Vous êtes sur la page 1sur 37

Personality and Social Psychology

Review
http://psr.sagepub.com/

A Meta-Analytic Review of Gender Variations in Adults' Language Use: Talkativeness, Affiliative Speech,
and Assertive Speech
Campbell Leaper and Melanie M. Ayres
Pers Soc Psychol Rev 2007 11: 328
DOI: 10.1177/1088868307302221
The online version of this article can be found at:
http://psr.sagepub.com/content/11/4/328

Published by:
http://www.sagepublications.com

On behalf of:

Society for Personality and Social Psychology

Additional services and information for Personality and Social Psychology Review can be found at:
Email Alerts: http://psr.sagepub.com/cgi/alerts
Subscriptions: http://psr.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://psr.sagepub.com/content/11/4/328.refs.html

>> Version of Record - Oct 12, 2007


What is This?

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

A Meta-Analytic Review of Gender Variations


in Adults Language Use: Talkativeness,
Affiliative Speech, and Assertive Speech
Campbell Leaper
Melanie M. Ayres
University of California Santa Cruz

Three separate sets of meta-analyses were conducted of


studies testing for gender differences in adults talkativeness, affiliative speech, and assertive speech. Across independent samples, statistically significant but negligible
average effects sizes were obtained with all three language
constructs: Contrary to the prediction, men were more
talkative (d = .14) than were women. As expected, men
used more assertive speech (d = .09), whereas women
used more affiliative speech (d = .12). In addition, 17
moderator variables were tested that included aspects of
the interactive context (e.g., familiarity, gender composition, activity), measurement qualities (e.g., operational
definition, observation length), and publication characteristics (e.g., author gender, publication source).
Depending on particular moderators, more meaningful
effect sizes (d > .2) occurred for each language construct.
In addition, the direction of some gender differences was
significantly reversed under particular conditions. The
results are interpreted in relation to social-constructionist,
socialization, and biological interpretations of genderrelated variations in social behavior.
Keywords:

communication; constructivism; conversation;


feminism; human sex differences; interpersonal
interaction; language; meta-analysis; sex roles;
socialization

he study of language in the construction and maintenance of gender divisions emerged as an active
research topic during the past three decades. In 1975, the
number of relevant studies was small enough for Thorne
and Henley (1975) to summarize all of them in an annotated bibliography. Interest in the relation between language and gender has since greatly expanded across the
social sciences. There have been narrative reviews of the

research in the fields of social psychology (e.g., Aries,


1996), developmental psychology (e.g., Leaper, 1994),
education (Swann, 1992), linguistics (e.g., Talbot, 1998),
communications (Dindia & Canary, 2006), sociology
(e.g., West & Zimmerman, 1985), anthropology (e.g.,
Maltz & Borker, 1982), and feminist studies (e.g.,
Crawford, 1995). A textbook on gender and language is
now in its seventh edition (J. T. Wood, 2007), and a
handbook on language and gender research was recently
published (Holmes & Meyerhoff, 2003). Moreover,
books aimed at the general public on gender and language
style have been best-sellers (e.g., Tannen, 1990).
Studying language may be especially fruitful when
examining ways that gender is negotiated and defined in
social interactions (Graddol & Swann, 1989). In general,
researchers have suggested that women are more likely
than men to use language to form and maintain connections with others (i.e., affiliation), whereas men are more
likely to use language to assert dominance and to achieve

Authors Note: The research was supported by grants to the first


author from the Academic Senate of the University of California,
Santa Cruz (No. 503035-19900) and the Social Sciences Division of
the University of California, Santa Cruz (No. 443060-09523). This
review started many years ago as part of a narrative review in the first
authors dissertation; he is grateful to Nancy Henley for her inspiration and guidance. In addition, Kris Anderson, Tera Hirsch, Catherine
Liguori, Tracy Marsh, Qhyrrae Michaelieu, Kelly Orrantia, Kathleen
Pettit, Paul Sanders, Sirinda Sincharoen, and Tara Smith are appreciated for their assistance at different phases of this work. Deborah
Tannen is thanked for her support. Judith Hall provided helpful comments. Address correspondence to Campbell Leaper, Department of
Psychology, University of California, Santa Cruz, CA 95064; e-mail:
cam@ucsc.edu
PSPR, Vol. 11 No. 4, November 2007 328-363
DOI: 10.1177/1088868307302221
2007 by the Society for Personality and Social Psychology, Inc.

328
Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


utilitarian goals (i.e., self-assertion).1 However, there is
considerable debate regarding the magnitude of observed
gender differences as well as the types of conditions under
which gender differences are most likely to occur. Metaanalytic reviews of the literature can be helpful in clarifying
overall trends and possible moderators.
Gender Differences in Language Use:
Talkativeness, Affiliative Speech, and
Assertive Speech
First, we will review the current literature on gender
differences in talkativeness, affiliative speech, and
assertive speech. Afterwards, we will contrast different
metatheoretical approaches that seek to explain gender
variations in language use. Finally, the possible influences of moderators will be considered.

Talkativeness
In our first meta-analysis, we test the popular stereotype that women are more talkative than men (see James
& Drakich, 1993). This view reflects the traditional
notions that women are expressive and affiliative whereas
men are stoic and independent. Three prior meta-analyses
provide some support for this premise. First, in their metaanalysis of gender effects on verbal ability across the lifespan, Hyde and Linn (1988) indicated a positive effect size
(d = .33) favoring females over males in tests of verbal
production. The studies in the meta-analysis were based
mostly on formal tests of language ability rather than
observations of actual conversations. Also, the authors
did not consider possible contextual moderators of gender
differences in verbal production.
Second, in a recent meta-analysis of gender-related
variations in childrens language use, Leaper and Smith
(2004) reported that girls tended to be more talkative
than did boys (d = .11). When the childs age was taken
into account, however, significant gender differences in
talkativeness were only seen among the 1- to 2 -yearolds (d = .32). This may have been due to the tendency
for language development to occur earlier in girls than
boys (Gleason & Ely, 2002). In addition, various contextual and methodological factors significantly moderated the likelihood and magnitude of gender differences.
The third meta-analysis suggesting that women may
be more talkative than men was by Leaper, Anderson,
and Sanders (1998). They examined studies comparing
mothers and fathers speech to their children and found
that mothers were more talkative with their children
than were fathers (d = .26). When aspects of the interactive context were taken into account, Leaper and
his co-authors found that differences between mothers
and fathers talkativeness occurred during relatively

329

unstructured activities not assigned by the researcher


(d = .30), but there was no parent gender difference during assigned problem-solving activities (d = .00). Thus,
the likelihood of gender differences in talkativeness may
depend on the situation.
In contrast to the previously reviewed meta-analyses,
an older meta-analysis (Hall, 1984) and a narrative
review (James & Drakich, 1993) reached a different conclusion about gender differences in talkativeness. They
deduced that most studies of adult conversation contradict the notion that women are more talkative than men.
Instead, relatively more studies indicated men were more
talkative than were women. Also, this pattern appeared
especially likely in mixed-gender groups pursuing specific
instrumental goals. Conversely, gender differences in
talkativeness appeared least likely during informal nontask-oriented contexts. However, the Leaper et al. (1998)
meta-analysis of parents speech indicated mothers were
more talkative than fathers during unstructured than
structured settings. Thus, both the relationship between
the conversational partners as well as the activity structure may influence the direction and the magnitude of
gender differences in language behavior. Possible moderators of gender differences in language use are more fully
addressed later in the introduction.

Affiliative and Assertive Speech


Besides differences in talkativeness, researchers have
described average differences in how women and
men use words to communicate. In this regard, we distinguish between affiliative and assertive speech.2 Affiliative language functions to affirm or positively engage
the other person. Examples include showing support,
expressing agreement, and acknowledging the others
contributions. Assertive language functions to advance
ones personal agency in a situation. It includes directive statements, giving information, and disagreeing
with or criticizing the others contributions. Average
gender differences in the uses of these language functions have been interpreted as manifestations of traditional gender divisions in society (e.g., Graddol &
Swann, 1989; Leaper & Smith, 2004). As proposed,
mens dominant status in society and their traditional
task orientation are enacted through their use of selfassertive language strategies such as directive and
instrumental speech. Womens relatively subordinate
status as well as their traditional caregiver role is
expressed through their use of affiliative language
strategies such as showing support and agreement.
Therefore, we tested if studies generally indicated, first,
greater use of affiliative speech among women than
men and, second, greater use of assertive speech among
men than women.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

330

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

Gender-related variations in affiliative and assertive


language have been highlighted in prior narrative
reviews of the research literature (e.g., Aries, 1996). In
addition, Leaper and Smiths (2004) meta-analysis of
childrens language found that on the average, girls used
more affiliative speech and boys used more assertive
speech. A parallel pattern emerged in Leaper et al.s
(1998) meta-analysis of parents speech to their
children: Mothers used more affiliative speech and less
assertive speech than did fathers. It is unclear, however,
if similar patterns would be seen in interactions between
adults. In one related meta-analysis, Roter, Hall, and
Aoki (2002) reported that female physicians were more
likely than male physicians to use affiliative communication (e.g., positive and emotion-focused talk). Other
reviewers have attempted to summarize the average
gender difference across studies. For instance, L. R.
Anderson and Blanchard (1982) indicated that men
tended to use certain forms of assertive communication
(active task behavior) more than did women, whereas
women tended to use some forms of affiliative communication (positive social-emotional behavior) more
than did men. However, the average difference for both
forms was less than 10% of total communicative acts.
Other reviewers have highlighted the role of situational
factors as moderators of gender differences in social
interaction (e.g., Aries, 1996, 1998; Deaux & Major,
1987; Eagly, 1987; LaFrance, Hecht, & Paluck, 2003;
Leaper et al., 1998; Leaper & Smith, 2004; Ridgeway
& Smith-Lovin, 1999). Before addressing the issue of
potential moderators in more depth, it will be helpful
to review different meta-theoretical explanations for
gender differences in social interaction.
Explanations for Gender Variations in Language Use
There are three major explanations for gender differences in language. They differ in how much weight is
assigned to socialization, situational demands, or biological predisposition. These explanations are often viewed
as complementary (e.g., see Leaper, 2000b; Leaper &
Friedman, 2007; Ruble, Martin, & Berenbaum, 2006;
W. Wood & Eagly, 2002). However, researchers tend
to adopt particular schools of thought that emphasize
the importance of one approach over others.
Two of these paradigms emphasize the influences of
social factors. The socialization explanation stresses the
cumulative influences of childrens participation in
gender-typed activities and their involvement in gendersegregated peer groups (e.g., see Leaper, 1994, 2000b;
Maccoby, 1998; Maltz & Borker, 1982; Thorne, 1993,
for reviews). According to this view, gender-segregated
peer groups lead girls and boys to establish and maintain
different norms, social identities, and preferences. This

includes the development of gender differences in affiliation and assertion. Girls affiliative concerns and behaviors are fostered through their greater participation in
dyadic interactions involving cooperative social-dramatic
activities, whereas boys self-assertive interests and behaviors are promoted through their greater participation in
solitary play as well as group interactions involving competitive or instrumental goals. Consequently, Maltz and
Borker (1982) proposed that girls learn to use their words
to create and maintain closeness with others through supportive and inclusive forms of speech. In contrast, the
authors argued that boys learn to use their words to
assert their position of dominance in relation to others
through commands and challenging statements. Maltz
and Borker further argued that gender differences in language use during childhood parallel observed gender
differences in adulthood.
A second type of explanation for gender differences
is the social constructionist (or contextualist) approach.
Researchers guided by this paradigm emphasize the
impact of interactive context rather than individual factors (Beall, 1993; Deaux & Major, 1987; Eagly, Wood,
& Diekman, 2000; Leaper, 2000b). According to the
social constructionist approach, women and men act
differently because of the demand characteristics of the
situation. One aspect of the situation considered especially important is mens greater status and power in
society (see Hall, 2006a; Henley, 1977, 2001). For
example, men may be more likely than women to dominate social interaction through higher rates of selfassertive speech; conversely, women may be more likely
to enact their traditionally subordinate status through
higher rates of affiliative speech. Accordingly, gender
differences in language behavior should be most likely
during mixed-gender interactions when this status effect
would be most salient (Carli, 1990). Carlis (1990)
research suggests this may be especially likely for gender
differences in assertive speech.
In addition, the activity setting is another aspect of the
context that may account for observed gender differences
in behavior. To the extent that women and men discuss
different topics or engage in different activities, there may
be corresponding differences in their language style (e.g.,
see Leaper et al., 1998). In other words, the context may
be partly driving what kind of language men and women
use. For example, affiliative speech may be more frequent
in self-disclosure tasks, whereas task-oriented activities
may occasion more assertive speech.
The biological explanation is the third type of explanation for gender differences in language use (see
Andersen, 2006; Hines, 2004). It is argued that sexrelated biological differences have resulted from evolutionary pressures for men to be more aggressive and
self-assertive and for women to be more nurturing and

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


affiliative (e.g., Andersen, 2006; Lippa, 2005; Luxen,
2005; Pellegrini & Archer, 2005). A premise underlying
this view is that women and men tend to differ in brain
organization and functioning. Indeed, some researchers
have proposed that womens brains have an advantage
for language ability (see Andersen, 2006). Support for
this view includes studies indicating a slight advantage
for girls over boys in language development and ability.
There are also some reports suggesting average sex differences in brain lateralization (e.g., see Gleason & Ely,
2002, for a review). However, the biological explanation
is mitigated by Hyde and Linns (1988) meta-analysis
indicating that the magnitude of gender differences
across measures of language ability was trivial (d = .11)
and that the magnitude of gender difference in language
abilities has decreased over the years.3 The biological
explanation is further complicated by some findings
that men tend to be more talkative than women (Hall,
1984; James & Drakich, 1993) as well as the variability
in gender differences in communication style across different situations. Thus, compared to social influences,
any biological bases for gender differences in overall
language ability may be negligible (Gleason & Ely,
2002) or mitigated by situational factors.
Finally, we wish to note that support for either the
socialization or the social constructionist explanations
in our analyses would not necessarily exclude possible
biological influences on gender-related variations in language use. In this regard, it is useful to distinguish
between strong and weak biological effects. If sexlinked biological factors account for a large proportion
of the variance in language use, then substantive gender
differences in language behavior should occur across
studies regardless of the year of study or aspects of the
interactive context. Few contemporary researchers
make such a claim. Instead, biologically oriented
researchers investigating gender-related social behaviors
generally acknowledge that first, biological predispositions can be altered over time through experience, and
second, existing dispositions can be mitigated or overridden by situational demands (e.g., Fitch & Bimonte,
2002; Hines, 2004; Lippa, 2005). These reflect relatively weak biological effects that were not amenable to
testing in the present meta-analyses.
Exploring Possible Moderators of Gender Differences
in Language Behavior
The tests for moderator variables are the most
important and interesting aspects of the meta-analyses.
We already know from narrative reviews that there are
inconsistent reports of gender differences in language use.
Whereas testing for average differences across studies
establishes how consistent and strong these trends

331

might be, it is more revealing to understand whether,


when, and where these differences occur. Toward this
larger goal, we tested 17 moderator variables. These
included operational definition, aspects of the interactive context, measurement qualities, and publication
characteristics.

Operational Definition
Students in research methods classes commonly learn
that how a construct is measured can affect the particular results one finds. Accordingly, we considered different
ways that researchers have operationally defined talkativeness, affiliative speech, and assertive speech. Some
measures of talkativeness, such as mean length of utterance (MLU) or the average duration of time spoken, may
be more sensitive to assessing the relative degrees that different speakers dominate the conversational floor.
There are also variations in specific types of affiliative and assertive speech. Some forms of affiliative
speech are simultaneously assertive, for example, as
when a person actively shows support or elaborates on
anothers comment (see Leaper, 1991; Penman, 1980).
In contrast, other forms of affiliative speech are relatively passive, for example, as when a person obligingly
goes along with the other. Next, there are types of
assertive speech that are domineering and emphasize the
persons power over the other as, for example, in the
use of commands. However, other forms of assertion
are less controlling, for example, as with many taskoriented speech acts such as giving information or making
suggestions (see Leaper, 1991; Leaper & Smith, 2004;
Penman, 1980). Thus, it is potentially interesting to see
if the likelihood of gender differences depends on the
specific type of affiliative or assertive speech.

Interactive Context
Several aspects of the interactive setting were examined
as possible moderators of gender influences on adults language behavior. Unfortunately, these did not include the
socioeconomic or cultural backgrounds of the research
participants. The available study samples were predominantly limited to those from middle-class European
American backgrounds. The contextual factors that we
did investigate included the participants undergraduate
status, the nature of the participants relationship to one
another, the number of persons in the group, the gender
composition of the group, whether the researcher was present during the observation, the physical setting, and the
activity. Each of these is reviewed below.
Undergraduate status. As often bemoaned in psychology research methods textbooks, our understanding
of human behavior is largely based on the study

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

332

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

of 19-year-old college students. This concern is pertinent to our meta-analyses because gender relations
among young college students may differ from those
among older adults. During college, students are often
exploring their identities as women and menand roles
for women and men as students are somewhat similar.
In contrast, older adults typically have established identities, and their roles as women and men are often differentiated in family and work. Hence, we contrasted
studies based on undergraduates versus older samples.

Researchers presence. In some studies, the researcher is


present while participants are interacting with one another.
This may heighten peoples self-consciousness and concerns with appearing in socially desirable wayswhich for
some people could include acting in gender-typed ways
(Deaux & Major, 1987). Therefore, gender differences
may be more likely when researchers are present while participants are talking. Alternatively, social desirability may
lead to the opposite effect; for example, men may act in a
more affiliative manner in front of a researcher.

Relationship. We expected that the participants relationship to one another would be an influential moderator. Prior studies suggest that gender differences are
more likely between strangers than between familiar
persons. With strangers, gender tends to act as a diffuse
status characteristic (W. Wood & Karten, 1986). Also,
people tend to be more concerned about presenting
themselves in socially desirable ways with strangers
(Deaux & Major, 1987). When there were a sufficient
number of studies, we also considered particular types
of relationships (e.g., friendships, spouses, or partners).

Setting. Earlier, we acknowledged the potential limitation of research samples based on only undergraduate
students. A related point is that many studies use university research laboratories to observe behavior. Relatively
fewer studies look at people in more naturalistic settings.
This may be an important factor, because research labs
are unfamiliar situations that can exaggerate peoples
gender-stereotyped behavior (see Deaux & Major, 1987).
Hence, we contrasted studies carried out in research labs
versus other settings.

Group size. The number of people participating in a


social interaction has been implicated as a potential
influence on gender-related differences in social behavior. Dyadic interactions tend to foster more intimacy
and reciprocity, whereas larger group interactions may
instill more competition (Bales & Borgatta, 1955;
Benenson, Nicholson, Waite, Roy, & Simpson, 2001;
Leaper, 1994). Therefore, group size was tested as a
moderator.
Gender composition. Numerous studies demonstrate
that gender-related variations in social behavior often
depend on the gender composition of the dyad or
group. Some gender differences may tend to emerge
when same-gender pairs or groups are compared,
whereas other differences may be more likely when
women and men are interacting with one another.
These two patterns are consistent with the socialization
and the constructionist interpretations, respectively (see
Carli, 1990). That is, the socialization explanation is
supported if the magnitude of any observed gender differences is stronger during same-gender interactions
than mixed-gender interactions. If women and men tend
to have different social norms, then participants presumably would be more likely to enact those norms
with partners from the same gender group. In contrast,
confirmation of the social constructionist view is indicated if gender differences are larger during mixedgender interactions. If gender differences are especially
likely in mixed-gender interactions, gender is implicated
as a status variable (W. Wood & Karten, 1986).

Activity. According to constructionist and contextualist models of gender, the activity is a highly influential
moderator of gender-related variations in social behavior (see Deaux & Major, 1987; Leaper, 2000a; Leaper
et al., 1998; Leaper & Smith, 2004). To consider activity as a moderator, distinctions were made between
studies observing unstructured interactions (e.g., leaving
participants in a room to discuss whatever they want)
and a variety of structured tasks (e.g., assigned discussion
topics, planning, toy play with children). It is unclear if
the assignment of gender-typed activities either increases
or decreases the magnitude of gender differences. Some
studies indicate that gender-typed activities increase participants concerns about acting in gender-typed ways
(e.g., see James & Drakich, 1993). In contrast, other
studies suggest that gender-related differences are actually mediated through the activityand that differences
in behavior are lessened when women and men are
engaged in similar activities (e.g., see Leaper et al., 1998).
The former pattern is more compatible with the socialization interpretation, whereasthe latter pattern is more consistent with the socialconstructionist interpretation.

Methodological factors
The meta-analysis also examined some potential
methodological factors that may moderate any gender
differences in language use. In addition to operational
definition, the observational method for recording
behavior and the length of observation were two measurement qualities that we examined. For observational
method, a distinction was made between audiotape,
videotape, and on-site coding. With videotape, the

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


researcher has the advantage of making repeated viewings as well as taking into account the nonverbal context. Therefore, one would expect the use of videotape
would be the most reliable method for measuring
behavior. However, Fagot and Hagan (1988) offered
evidence suggesting live observation may be more accurate than videotape under certain conditions. The metaanalysis will help us test if either of the recording
methods is more reliably associated with larger effect
sizes. In addition, the length of behavioral observation
is another methodological feature that may influence
the reliability and the validity of the results. Consistent
behavior patterns should be more apparent with longer
observation periods (see Fagot, 1985).
We additionally took into account four publication
characteristics. First, this included the first authors
gender. In some prior meta-analyses, the magnitude and
the direction of gender differences in social behavior varied depending on the first authors gender (e.g., K. J.
Anderson & Leaper, 1998; Eagly & Carli, 1981; Leaper
et al., 1998). Second, we contrasted whether the studys
primary research question concerned gender. Because
gender is routinely included as a factor in many statistical
designs, it was also possible to test for possible bias
toward reporting significant results in studies designed
specifically to investigate gender effects on behavior.
Third, publication source was tested as a moderator. We
compared studies published in top-tier journals versus
other sources as a rough way to see if publication quality
influenced the likelihood of gender effects. Finally, year
of study was used to explore if gender differences in language behavior have possibly declined over the years.

Summary
Two general questions guided the present study:
First, to what extent do women and men differ in their
language use? Research suggests there are average
gender differences in talkativeness, affiliative speech,
and assertive speech. We tested the hypotheses that
women would be more talkative, use more affiliative
speech, and use less assertive speech than would men.
Second, what types of factors moderate the incidence
and magnitude of any observed gender effects? To this
end, we tested several contextual and methodological
moderator variables. We expected that the magnitude
of average gender differences would be negligible and
that more meaningful effects would occur when the
moderators were taken into account.

METHOD
Literature Search
Studies examining gender-related effects on adults
talkativeness, affiliative language, or assertive language

333

(defined below) were collected through a variety of


sources. Most of the studies were identified through
searches of the PsycInfo database. Studies were also
identified through citations in these papers as well as
various review articles and books. The dates of publication for the collected studies ranged from 1968 to
2004.
Three selection criteria were used: First, only studies
that tested for gender effects on adults language behavior were used. Second, only studies using quantitative
observational measures were included. Therefore, selfreport studies of verbal behaviors were excluded. Also,
researchers global ratings of communication style were
not used. Third, only studies published in either
research journals or books were included.
Language Variables
Three separate sets of meta-analyses were performed
for studies of gender differences in talkativeness, affiliative speech, and assertive speech, respectively. Intercoder
reliability was assessed between two researchers for classification of the specific operational definitions for each
language variable ( = .84). In each meta-analysis, all of
the language measures were based on frequency, proportion, or rate scores. Each language variable is further
described below.
Amount of talking. Among those studies testing for
gender differences in talkativeness, there were 63 published studies. A distinction was made between the following operational definitions of amount of talking: (a)
number of words or utterances, (b) rate or time sampling,
(c) MLU or words per turn, (d) duration of talking, (e)
total turns, and (f) total statements or speech acts. In
addition, there was one study (Pillon, Degauquier, &
Duquesne, 1992) that reported a MANOVA combining
MLU, total words, and total turns.
Affiliative speech. Affiliative speech refers to verbal
acts that affirm the speakers connection to the listener.
There were 47 published studies testing for gender differences in affiliative language. A distinction was made
between the following types of affiliative speech: (a)
supportive (e.g., praise, approval, collaboration), (b)
active understanding (reflective comments, probing
questions), (c) agreement, (d) acknowledgment (including minimal listening responses), and (e) general socioemotional speech (e.g., a combination of expressing
solidarity, affection, and support). Researchers using
the latter category were using Baless (1970) scheme (or
one similar to it).
Assertive speech. Assertive speech refers to verbal
acts that seek to influence the listener. There were
39 published studies identified that tested for gender

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

334

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

differences in assertive language. The following forms of


assertive speech were coded: (a) directive (imperative
statements or direct suggestions), (b) giving information
(descriptive statements or explanations), (c) suggestions
(suggestions, problem solving, or giving opinion), (d) criticism (criticism or disapproval), (e) disagreement, and
(f) general task-oriented speech (e.g., a combination of giving suggestions, opinions, or direction). Researchers using
the latter were typically using Baless (1970) categories (or
similar). In addition, there were two samples that did not
fit into any of the categories. In one of these samples (Reid
et al., 2003), a measure of tentative speech was used. (For
the meta-analysis, the hypothesized direction was for men
to score lower than women.) In the other sample (Scudder
& Andrew, 1995), a measure of verbal threat was used.
According to our conceptual model (Leaper, 1991;
Penman, 1980), affiliation and assertion are two dimensions underlying speech acts; that is, they are not mutually exclusive. Active understanding, verbal support, and
general socioemotional speech are affiliative categories
that reflect high affiliation and high assertion (i.e., their
aim is both to affirm and to influence the listener),
whereas agreement and acknowledgment reflect high
affiliation and relatively low assertion (i.e., they affirm
other but downplay ones influence). All of the assertive
categories are relatively low to moderate affiliation.
Directives, criticism, and disagreement are low in affiliation because they tend to distance the speaker from the
listener. Offering suggestions and general task-oriented
speech are moderately affiliative because they generally
engage the listener. Informing speech is moderately
assertive inasmuch that giving information can influence
another person; it is also moderately affiliative to the
extent that it maintains a connection with the listener.
Other Moderator Variables
In addition to investigating the magnitude of gender
effects associated with the different language behaviors,
several moderator variables were examined. Each of these
factors is summarized below. Also, the characteristics for
each moderator variable associated with each study are
presented for each meta-analysis in Tables 1, 2, and 3.
In some studies, there was inadequate information to
determine the values of some moderator values. Despite
efforts to track down the missing information (e.g., contacting author), some of the moderators have missing values for a few studies. The respective numbers of these
cases for each moderator are noted in the Results section.
Features of the Interactive Context
The available study samples were predominantly limited to participants with middle-class, European
American backgrounds. Our analysis of contextual moderators was limited to the following seven factors: (1) the

participants undergraduate status (yes or no), (2) the


relationship between the participants (strangers, classmates, friends, dating partners, spouses or cohabiting
partners, parents with their child or children present),
(3) the size of the group being observed (dyads vs. larger
groups), (4) the gender composition of the group (same
or mixed gender), (5) the researchers presence (no or yes)
during the observed interaction, (6) the observational setting (lab or other setting), and (7) the activity in which
participants were engaged while being observed.
Activities were classified as either (a) unstructured
(i.e., participants could discuss whatever they wanted),
(b) an assigned self-disclosure discussion, (c) an assigned
discussion about a personal disagreement, (d) an
assigned discussion about a nonpersonal topic (e.g., a
current news event), (e) an assigned deliberation (e.g.,
decision-making task or a debate), (f) a naturalistic task
(e.g., observations at a public information booth), (g) a
game or a test, (h) play with a child, (i) a nonplay interaction with a child, or (j) mixed or (k) unclear activities.
Intercoder agreement for these 10 activity codes was
acceptable ( = .81). However, due to low occurrences,
the following activity categories were combined:
Naturalistic tasks and games/tests were combined to
create a miscellaneous task category. Also, play and
nonplay interactions with children were combined.
Methodological Characteristics
In addition to operational definition, other methodological moderators that were tested included (1)
method for recording behavior (audiotape, videotape,
or on-site coding), (2) length of observation, (3) first
authors gender, (4) gender focus of the study, (5) publication year, and (6) publication source. For observation length, we contrasted those that were very brief
(less than 10 min), somewhat short (less than 20 min),
and somewhat long (at least 20 min). (There were relatively few studies that involved observation lengths
much longer than 20-30 min; exploratory analyses indicated that making further distinctions was not helpful.)
The gender-focus moderator refers to whether the
studys primary research question addressed gender as a
topic. Although most collected studies addressed
gender-related variations in social behavior, some studies were not primarily concerned with gender per se but
included it as an exploratory or control factor in their
analyses. For publication year, we made an approximate median split and contrasted studies published
before 1986 and those published in 1986 or later. When
testing publication source as a moderator, a distinction
was made between studies published in a top-tier
journal (e.g., American Psychological Association
journals such as Journal of Personality & Social
Psychology) and other sources (including books and
other journals). Although many excellent studies are

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

(text continues on page 344)

335

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Argyle, Lalljee, &


Cook (1968)
Aries (1982)
Athenstaedt,
Haas, &
Schwab (2004)
Athenstaedt et al.
(2004)
Bilous & Kraus
(1988)a
Bilous & Kraus
(1988)a
Brooks (1982)
Brouwer, Gerritsen,
& DeHaan
(1979)
Brown (1990)
Campbell, Kleim,
& Olson (1992)b
Campbell et al.
(1992)b
Carli (1990)
Case (1988)
Cashdan (1998)c
Cashdan (1998)c
Chelune (1976)
Cherulnik (1979)
Craig & Pitts (1990)
Crosby, Jose, &
Wong-McCarthy (1981)
Dabbs & Ruback (1984)
Davis & Gilbert (1989)
Dovidio, Heltman,
Brown, Ellyson, &
Keating (1988)
Frances (1979)
Gall, Hobby, &
Craik (1969)
Hammen & Peplau
(1978)
Hannah & Murachver
(1999)
Hawkins & Power
(1999)
Heiss (1962)
Hladik & Edwards (1984)b
16
65

30
30
15
30
72

298
16
43
43
59
5
29
29
12
18
8
48
10
60

24
44
20
37
32
39
54
10

30
38
15
30
222

309
16
61
61
59
5
47
49
12
18
8
48
10
60

24
44
19
37
32
59
54
10

N (m)

16
53

N (w)

Trn
TW
TW

Dur

Rat

TW

Dur
Dur

TW
Dur
Dur

Dur
TW
TW
Dur
Dur
TW
Dur
Dur

TW

TW
Dur

TW
Dur

TW

Dur

Dur

Rat
Trn

Op Def

0.70
0.326
0

p = .001
p = .01
p = .50

0
1.15
0.64

p = .50
t = 2.58
F = 12.15

0.57
0
0
0.17
0.61
0.87
0
0

t = 2.84
p = .50
p = .50
t = 0.74
t = 0.61
F = 4.5
p = .50
p = .50

p = .50

0.60

t = 2.99

0
0.62

p = .50
F = 6.07

p = .50

0.18
0.27

t = 0.98
G

0.349

2.04

t = 5.59

t = 1.09

0.35

F = 2.08

0.90
0.543

0.38

F = 4.32

G
F = 6.49

0.61
0.39

p = .05
p = .02

Stat Value

Study Characteristics for Talkativeness Meta-Analysis

Author and Year

TABLE 1:

W
M
W

M
W

W
M
W

M
W
W
W
W
M
M
U

W
M

U
W

M
W

Auth

Oth
Oth
Oth

Oth

Oth

Oth

Top
Oth

Oth
Oth
Top

Oth
Top
Oth
Oth
Oth
Oth
Oth
Oth

Oth

Oth
Oth

Oth
Oth

Oth

Oth

Oth

Oth
Oth

Pub

Yes
Yes
Yes

Yes

Yes

Yes

Yes
Yes

Yes
Yes
Yes

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes

Yes
Yes

Yes
Yes

Yes

Yes

Yes

No
Yes

UG
UG
Par

UG

UG

UG

UG
Oth

UG
UG
UG

UG
UG
Oth
UG
UG
UG
UG
UG

UG

Oth
UG

UG
Oth

UG

UG

UG

UG
UG

Gender Pop

Oth
Dat
C/S

Res

Str

Res

Str
Str

Str
Str
Str

Str
Str
Oth
Frn
Oth
Res
Res
Oth

Str

Str
Str

Str
Oth

Str

Frn

Frn

Str
Str

Rel

Both
MG
MG

Both

Both

Both

MG
Both

MG
SG
MG

MG
Both
MG
SG
MG
N/A
Both
MG

MG

Both
MG

SG
MG

SG

SG

MG

MG
Both

Comp

Lab

Lab

Lab

Lab
Lab

5
2
3

2
2

2
5
2

6
2
10
10
8
2
2
5

2
2

Lab
Lab
Oth

Lab

Oth

Lab

Lab
Lab

Lab
Lab
Lab

Lab
Lab
Oth
Lab
Lab
Lab
Lab
Oth

Lab

Oth
Lab

No

No

No

No
No

Res

30
20
30

12

18

9
14

6
20
7

N/A
10
45
15
15
15
N/A
40

N/A

3
6

No
No
No

No

No

Yes

No
No

No
No
No

N/A
No
No
No
No
Yes
Yes
No

N/A

No
No

10
No
2,880 Yes

10

20

20

15
10

Setting Long

2
Lab
N/A Oth

2
5

Size

A
A
A

V
V

A
V
A

V
V
A
V
V
A
V
A

A
V

A
A

A
V

Rec

(Continued)

A-D
A-D
CH

M-T

Unst

A-T

A-T
Unst

A-D
Unst
A-D

Unc
A-D
N-T
A-D
A-D
SD
SD
N-T

Unc

N-T
A-T

A-D
N-T

A-D

A-D

A-D

A-T
A-D

Activity

336

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

10
70
25
34
64
72
72
15
76
28
39
30
48
20
20
5
22
22

12
12
20
20

20
54
54
36

96

10
71
25
37
64
72
72
15
76
28
39
30
48
20
20
5
22
22

12
12
27
27

20
54
54
36

72

Hladik & Edwards (1984)b


C. Johnson (1994)
Jones, Gallois, Callan, &
Barker (1995)
Kelly, Wildman, &
Urey (1982)
Kimble & Musgrove
(1988)
Kimble, Yoshikawa,
& Zehr (1981)b
Kimble et al. (1981)b
Kollock, Blumstein, &
Schwartz (1985)
Leaper (1987)
Leffler, Gillespie, &
Conaty (1982)
Margolin & Wampold
(1981)
Markel, Long, &
Saine (1976)
Marlatt (1970)
Martin & Craig (1983)b
Martin & Craig (1983)b
Martin, Davis, &
Dancer (1996)
McLaughlin (1991)b
McLaughlin (1991)b
McMillan, Clifton,
McGrath, & Gale
(1977)a
McMillan et al.
(1977)a
Mehl & Pennebaker
(2003)
Mehl & Pennebaker
(2003)
Moore, Shaffer,
Goodsell, &
Baringoldz (1983)
Mulac (1989)a
Mulac (1989)a
Natale, Elliot, &
Jaffe (1979)
Nemeth, Endicott,
J., & Wachtler
(1976)

N (m)

N (w)

(Continued)

Author and Year

TABLE 1:

Sta

Dur

MU
Dur
Dur

MU

TW

Sta

Sta

Trn
Dur
Rat

Dur
Rat
TW
MU

Rat

Dur

Dur
MU

Dur
Rat

Dur

Dur

Dur

MU
Dur

Op Def

0
0
0.46
0
0
0.91

p = .50
p = .50
F = 3.2
p = .50
p = .50
F = 1.42

0.60

0
0.03
0.35
0.59

0.27

p = .50
M and SD
M and SD
r = .28
F = 3.02

0.18

t = 0.63

p = .05

0.12

t = 0.29

0.80
0
0

p = .049
p = .50
p = .50

p = .5

0.69
0.62
0
0

7.15
9.32
.50
.50

=
=
=
=

0.36

F = 4.19

F
F
p
p

0.84

0.86

F = 9.17
G

0.20
0

M and SD
p = .50

Stat Value

M
M
M

W
M
M

M
M
W
W

M
M

M
M

W
W

Auth

Oth

Top

Oth
Oth
Oth

Top

Top

Oth

Oth

Oth
Oth
Oth

Oth
Top
Oth
Oth

Top

Oth

Oth
Oth

Top
Top

Oth

Oth

Oth

Oth
Oth

Pub

Yes

No

Yes
Yes
Yes

No

No

Yes

Yes

Yes
Yes
Yes

Yes
No
Yes
Yes

Yes

Yes

Yes
Yes

Yes
Yes

Yes

Yes

Yes

Yes
Yes

UG

UG

UG
UG
UG

UG

UG

UG

UG

Oth
UG
UG

UG
UG
UG
UG

Oth

UG

Oth
UG

UG
UG

UG

UG

Oth

Par
UG

Gender Pop

Str

Str

Str
Oth
Oth

Str

Str

Str

Str

Oth
Oth
Oth

Str
Res
Str
Str

Spo

Str

Spo
Str

Str
Str

Str

Str

Str

C/S
Str

Rel

MG

Both

Both
SG
MG

Both

Both

MG

SG

MG
Both
Both

Both
Both
Both
Both

MG

Both

MG
MG

Both
Both

MG

MG

SG

MG
Both

Comp

Lab

Lab

N/A
Lab
Lab

Lab
Lab
Lab
Lab

Lab

Lab

Oth
Lab

Lab
Lab

Lab

Lab

Lab

Oth
Lab

2
2
2

Lab

Lab

Lab
Lab
Lab

N/A Oth

120

30

N/A
8
8

N/A

N/A

30

30

5
N/A
N/A

N/A
5
4
4

20

30

N/A
5

10
10

N/A

30
10

Setting Long

N/A Oth

2
2
2

2
2
2
2

2
2

4
4

3
3

Size

No

No

No
No
No

No

No

No

No

No
No
No

No
Yes
No
No

No

No

No
No

No
No

No

No

No

No
No

Res

A
V
V

V
V
V

A
A
V
V

A
A

A
A

A
V

Rec

A-D

Unst

Test
A-D
A-D

Unst

Unst

Test

Test

Mix
A-D
A-D

Unst
Unst
Unst
Unst

DS

M-T

A-D
A-D

Mix
Mix

A-D

A-D

A-T

CH
M-T

Activity

337

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

20
57

47
21
18
20
24
24
43
20
14
18
13
13

38
40
10
20
18
18
16

20
57

50
21
18
20
22
22
57
20
14
18
13
13

38
40
10
20
18
18
16

TW
TW
MU
MU
Sta

Rat

TW

Dur

Dur

Rat
TW
Dur
Trn

Dur
TW
MU

Dur
Rat

Dur

Dur

Oth

0.52

p = .01
0.28
1.0
2.14
1.60
0.59

p = .50

M and SD
M and SD
t = 6.43
t = 6.76
F = 5.63

0.78

0
0
0.7
0

p = .01

.50
.50
3.47
.50
0

=
=
=
=

p = .50

p
p
F
p

0.16
0
0.69

t = 0.73
p = .50
F = 5.4

0.49

r = 0.24
0.36
0.09

0.97

F = 26.59

G
F = 0.14

0.27

F = 1.5

W
W
W
W
W

M
M
W
M

W
M
M

M
W

Oth
Oth
Oth
Oth
Top

Top

Oth

Oth

Oth

Top
Oth
Oth
Oth

Oth
Top
Top

Oth
Oth

Oth

Top

Oth

Yes
Yes
Yes
Yes
Yes

Yes

Yes

Yes

Yes

No
Yes
Yes
Yes

Yes
No
No

Yes
No

No

Yes

Yes

UG
UG
UG
UG
UG

Par

UG

UG

UG

UG
Oth
Oth
Oth

Oth
UG
UG

UG
Oth

Oth

UG

UG

Str
Str
Str
Str
Str

C/S

Str

Str

Str

Str
Res
Str
Oth

Spo
Str
Str

Str
Spo

Str

Str

Str

Both
Both
SG
MG
MG

MG

MG

MG

SG

MG
Both
MG
MG

MG
SG
SG

MG
MG

SG

MG

MG

2
2
2
2
4

5
2
4
2

2
2
2

2
2

Lab
Lab
Lab
Lab
Lab

Oth

Lab

Lab

Lab

Lab
Oth
Lab
Lab

Oth
Lab
Lab

Lab
Lab

Lab

Lab

Lab

5
5
60
60
5

60

15

15

20
N/A
16
20

7
5
5

10
7

10

10

10

No
No
Yes
Yes
No

Yes

No

No

No

No
Yes
No
No

No
No
No

No
No

No

No

No

V
V
A
A
V

V
A
V
A

A
A
A

V
V

A-D
A-D
A-T
A-T
A-D

CH

Unst

A-D

A-D

A-D
M-T
A-D
A-D

Unst
Unc
Unc

A-D
DS

Unst

A-D

Mix

NOTE: N (w) = number of women; N (m) = number of men; Op Def = operational definition (Dur = duration; MU = mean length of utterance; Rat = rate or time sampling; Sta = total
statements; Trn = total turns; TW = total words; Oth = other); Stat value = statistical value (G = Effect size computed based on aggregate of similar measures for same operational definition; M and SD = effect size computed from reported means, standard deviations, and group sizes); G = index of effect size; Auth = first authors gender (M = man; W = woman;
U = unclear); Pub = publication source (Top = top journal; Oth = other source); Gender = gender study (yes or no); Pop = type of population sampled (UG = undergraduate students;
Par = parents; Oth = other); Rel = relationship between participants (C/S = participant with own child and spouse; Dat = dating couple; Frn = friends; Res = researcher and participant;
Spo = spouses or domestic partners; Str = strangers; Oth = other); Comp = gender composition of group (Both = same gender and mixed gender combined; MG = mixed gender; SG = same
gender; Oth = other); Size = group size (number of participants observed interacting together); Setting = observational setting (lab or other); Long = length of observation (in min); Res =
whether the researcher was present during the observation (no or yes); Rec = method of recording interaction (A = audiotape only; O = on-site coding only; V = videotape); Activity = type
of activity observed (A-D = assigned decision making or debate topic; A-T = assigned general topic(s); CH = interaction with child; DS = discussion about a disagreement; M-T = miscellaneous or mixed task activity; Mix = mixed activities; N-T = naturalistic task-oriented activity; SD = self-disclosure; Test = assigned test or puzzle; Unc = unclear; Unst = unstructured
activity); N/A = information not available.
a. Reported separate tests of gender differences in talkativeness in two independent samples.
b. Reported separate tests of gender differences in talkativeness using two different operational definitions with the same sample.
c. Reported tests of gender differences in talkativeness from two different studies with independent samples.

Pillon, Degauquier,
C., & Duquesne
(1992)
Porter, Geis, Cooper,
& Newman (1985)
Pushkar, Basevitz,
Arbuckle,
Nohara-LeClair,
Lapidus, & Peled
(2000)
Reid, Keerie, &
Palomares (2003)
Resick et al. (1981)
Robey, Canary, &
Burggraf (1998)
Rosenfeld (1966)b
Rosenfeld (1966)b
Ruback, Dabbs, &
Hopper (1984)
Rubin & Nelson (1983)
Schmid-Mast (2001)
Shaw & Sadler (1965)
Simkins-Bullock &
Wildman (1991)a
Simkins-Bullock &
Wildman (1991)a
Stiles, Walz, Schroeder,
Williams, & Ickes
(1996)
Stuckey, McGhee,
& Bell (1982)
Turner, Tjaden, &
Weismer (1995)
Turner et al. (1995)
M. Wood (1966)a
M. Wood (1966)a
W. Wood & Karten (1986)

338

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

10
16
16
16
16
32
32
32
32
32
32
32
32
5

48
50
24
24
44
84
84
32
79
39
54
70
20
25
25
16
16
17
17
34
15

N (w)
16
16
16
16
16
32
32
32
32
32
32
32
32
5

48
54
24
24
44
84
84
32
79
59
54
71
20
25
25
22
22
33
33
37
15

Caldwell & Peplau (1982)


Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989) a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Case (1988)
Crosby, Jose,
& WongMcCarthy (1981)
Dixon & Foster
(1998)
Eskilson & Wiley
(1976)
Fagot (1978)
Frances (1979)
Grotevant &
Cooper (1985)b
Grotevant &
Cooper (1985)b
Hannah &
Murachver (1999)
Hauser et al. (1987)
Hawkins & Power (1999)
Heiss (1962)
C. Johnson (1994)
C. Johnson, Funk, &
Clay-Warner (1998)
Jones, Gallois, Callan,
& Barker (1995)b
Jones et al. (1995)b
Julien et al. (2000)c
Julien et al. (2000)c
Julien et al. (2000)c
Julien et al. (2000)c
Kelly, Wildman, &
Urey (1982)
Kollock, Blumstein, &
Schwartz (1985)

N (m)

Ack

Sup

Agr
Ack
Gen
Ack
Gen
Ack

Gen

Ack
Sup
AU
Gen
Ack

Ack

Agr

Gen
Sup
Ack

Ack

Ack

Sup
Gen
Gen
Agr
Agr
Gen
Gen
Agr
Agr
Gen
Gen
Agr
Agr
Agr

Op Def

Study Characteristics for Affiliative Language

Author and Year

TABLE 2:

0.70
0

G
p = .5

0.85
0.75
0.47
0.01
0.12
0.19

F = 9.04
F = 7.04
M and SD
M and SD
M and SD
M and SD

0
0.15
0.44
0.33
0

p = .5
F = 1.82
2 = 4.28
p = .01
p = .5
0.39

p = .5

M and SD

0.21

F = 1.10

p = .5

0.29

F = 3.95

0.90
0
0

0.74
1.79
1.50
0.53
0.91
0.18
0.35
0.16
0.19
1.26
1.16
0
0
0.63

t = 1.84
t = 5.05
t = 4.23
t = 1.51
T = 2.56
t = 1.04
t = 1.98
t = 0.88
t = 1.06
t = 5.01
t = 4.63
p = .5
p = .5
p = .09

G
p = .5
p = .5

Stat Value

W
W
W
W
W
W

W
M
W
M
W

W
W
W

W
W
W
W
W
W
W
W
W
W
W
W
W
W

Auth

Oth

Oth

Oth
Oth
Top
Top
Top
Top

Oth

Oth
Oth
Oth
Oth
Oth

Top

Top

Oth
Top
Oth

Oth

Oth

Oth
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Oth

Pub

Yes

Yes

Yes
Yes
Yes
Yes
Yes
Yes

Yes

Yes
Yes
Yes
Yes
Yes

No

No

Yes
Yes
Yes

Yes

Yes

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Oth

UG

Oth
Oth
Oth
Oth
Oth
Oth

UG

UG
Par
UG
UG
UG

Par

Par

UG
Par
Oth

UG

UG

UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
Oth

Gender Pop

Spo

Str

Str
Str
Frn
Frn
Frn
Frn

Str

Res
C/S
Oth
Dat
Str

C/S

C/S

Str
C/S
Str

Str

Str

Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Oth

Rel

MG

MG

SG
SG
SG
SG
SG
SG

SG

Both
Both
Both
MG
Both

MG

MG

Both
MG
Both

Both

MG

SG
SG
SG
SG
SG
MG
MG
MG
MG
Both
Both
Both
Both
MG

Comp

2
2
2
2
2
2

2
3
5
2
3

3
3
2

2
2
2
2
2
2
2
2
2
2
2
2
2
10

Size

Oth

Lab

Lab
Lab
Lab
Lab
Lab
Lab

Lab

Lab
Lab
Lab
Lab
Lab

Lab

Lab

Lab
Oth
Lab

Lab

Lab

Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Oth

N/A

N/A

8
8
20
20
20
20

11

12
N/A
30
20
10

20

20

N/A
300
14

6
20
20
20
20
20
20
20
20
20
20
20
20
45

Setting Long

No

No

No
No
No
No
No
No

No

No
No
No
No
No

No

No

No
Yes
No

No

No

No
No
No
No
No
No
No
No
No
No
No
No
No
No

Res

V
V
V
V
V
V

V
A
A
A
V

V
O
V

A
V
V
V
V
V
V
V
V
V
V
V
V
A

Rec

A-D

A-D

A-T
A-T
DS
DS
DS
DS

A-D

M-T
A-D
A-D
A-D
M-T

A-D

A-D

Test
CH
Unst

M-T

A-D

Unc
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
N-T

Activity

339

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Carson, Baker,


Holliday, & Myers
(1995)d
Leaper et al. (1995)d
Leaper et al. (1995)d
Leaper et al. (1995)d
Leaper (1998)e
Leaper (1998)e
Leik (1963)e
Leik (1963)e
Margolin & Wampold
(1981)b
Margolin & Wampold
(1981)b
Martin & Craig (1983)
McLaughlin (1991)
McLaughlin, Cody, Kane,
& Robey (1981)b
McLaughlin et al. (1981)b
McMullen, Vernon, &
Murton (1995)f
McMullen et al. (1995)f
McMullen et al. (1995)f
McMullen et al. (1995)f
Moore, Shaffer, Goodsell,
& Baringoldz (1983)
Natale, Elliot, & Jaffe
(1979)
Piliavin & Martin (1978)b
Piliavin & Martin (1978)b
Pillon, Degauquier, &
Duquesne (1992)b
Pillon et al. (1992)b
Reid, Keerie, &
Palomares (2003)
Robey, Canary, &
Burggraf (1998)
Roger & Nesshoever
(1987)
Roger & Schumacher
(1983)
Roopnarine & Adams
(1987)
Sayers & Baucom (1991)
Simkins-Bullock &
Wildman (1991)e
18
18
18
18
19
24
9
9
39
39
20
22
60
60
20
10
17
17
20
36
76
76
20
20
42
20
28
36
37
60
13

18
18
18
18
25
24
9
9
39
39
20
22
60
60
20
10
17
17
20
36
76
76
20
20
42
20
28
36
37
60
13

Sup

Sup
Gen

Ack

Ack

Ack

Ack

Gen
Ack

Ack
Gen
Agr

Sup

Ack
Ack
Ack
Ack

Sup
Ack

Agr
Ack
Ack

Sup

Ack
AU
Ack
AU
Agr
Agr
Gen
Gen

0.32
0.70
0.74
0
0.06

t = 1.45
F = 6.89
F = 9.88
p = .5
M and SD

p = .5

p = .5

0
0

0.08
0.62
0.52

p = .5
p = .5

r = .04
F = 29.52
F = 20.78

p = .5

0.04
0.76
0.19
0.50

0.78
0.14

F = 18.15
F = 1.23
1.69
1.69
0.19
2.08

0.05
0
0

F = 0.09
p = .5
p = .5

=
=
=
=

0.32

F = 3.95

t
t
t
t

0
0.92
0
0.08
0.47
0.28
0.37
0.13

M and SD
M and SD
M and SD
M and SD
F = 2.37
F = 1.88
PR
PR

M
M

W
W

M
W
W

W
W
W
W

W
W

W
W
M

M
M
M
M
M
M
M
M

Oth

Oth
Top

Top

Oth

Oth

Oth

Oth
Oth

Top
Oth
Oth

Oth

Oth
Oth
Oth
Oth

Oth
Oth

Top
Oth
Oth

Top

Oth
Oth
Oth
Oth
Oth
Oth
Oth
Oth

Yes

No
Yes

Yes

Yes

Yes

Yes

Yes
Yes

No
Yes
Yes

Yes

Yes
Yes
Yes
Yes

Yes
Yes

Yes
Yes
Yes

Yes

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

UG

Par
Oth

UG

UG

Oth

UG

UG
UG

UG
UG
UG

UG

UG
UG
Oth
Oth

UG
UG

Oth
UG
UG

Oth

UG
UG
UG
UG
UG
UG
Par
Par

Str

C/S
Spo

Str

Str

Spo

Str

Str
Str

Str
Str
Str

Str

Str
Str
Spo
Spo

Str
Str

Spo
Str
Oth

Spo

Frn
Frn
Frn
Frn
Frn
Frn
Str
C/S

SG

MG
MG

SG

MG

MG

MG

MG
MG

Both
Both
Both

Both

SG
MG
MG
MG

Both
Both

MG
Both
Both

MG

SG
SG
MG
MG
SG
MG
MG
MG

3
2

2
2

2
4
4

2
2
2
2

2
2

2
2
2

2
2
2
2
2
2
3
3

Lab

Lab
Lab

Lab

Lab

Lab

Lab

Lab
Lab

Lab
Lab
Lab

Lab

Lab
Lab
Oth
Oth

Lab
Lab

Lab
Lab
Lab

Lab

Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab

15

8
N/A

10

10
10

30
35
35

N/A

30
30
30
30

30
30

20
4
N/A

20

10
10
10
10
10
10
N/A
N/A

No

No
No

No

No

No

No

No
No

No
No
No

No

No
No
No
No

N/A
N/A

No
No
No

No

No
No
No
No
No
No
No
No

V
V

A
A

A
A
A

A
A
V
V

A
A

O
V
V

A
A
A
A
A
A
O
O

(Continued)

A-D

CH
DS

A-D

A-D

Unst

A-D

Mix
Mix

Unst
A-T
A-T

Test

Unst
Unst
Unst
Unst

Mix
Mix

DS
Unst
A-D

DS

SD
SD
SD
SD
A-D
A-D
A-D
A-D

340

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

(Continued)

13
278
278

38
38
19
10
20
16
28
32

38
38
19
10
20
16
28
32

N (m)

13
278
278

N (w)

Gen
Gen

Gen

Agr
Agr

Agr
Ack
Gen

Sup
Agr
Ack

Op Def

0.67
0
0.48

G
p = .5
F = 7.4

0.31
0.78

0
0
0.15

p = .5
p = .5
t = 0.65
G
G

0
0.10
0

p = .5
p = .05
p = .5

Stat Value

W
W

M
M

M
M
M

W
W
W

Auth

Oth
Oth

Top

Top
Top

Oth
Oth
Oth

Oth
Oth
Oth

Pub

Yes
Yes

Yes

Yes
Yes

Yes
Yes
Yes

Yes
Yes
Yes

UG
UG

UG

UG
UG

UG
UG
UG

UG
Oth
Oth

Gender Pop

Str
Str

Str

Str
Str

Str
Str
Str

Str
Spo
Spo

Rel

SG
MG

MG

Both
Both

MG
MG
MG

MG
MG
MG

Comp

2
2

2
2

2
2
2

2
2
2

Size

Lab
Lab

Lab

Lab
Lab

Lab
Lab
Lab

Lab
Lab
Lab

N/A
N/A

5
5

6
6
6

15
15
15

Setting Long

No
No

No

No
No

No
No
No

No
No
No

Res

V
V

V
V

V
V
V

A
V
V

Rec

Unc
Unc

A-D

A-D
A-D

Unst
Unst
Unst

A-D
DS
DS

Activity

NOTE: N (w) = number of women; N (m) = number of men; Op Def = operational definition (Ack = acknowledge or minimal listening response; Agr = agreement; AU = active understanding; Gen = general affiliation; Sup = support, praise, or approval); Stat value = statistical value (G = effect size computed based on aggregate of similar measures for same operational definition; M and SD = effect size computed from reported means and standard deviations); G = index of effect size; Auth = first authors gender (M = man; W = woman); Pub = publication source
(Top = top journal; Oth = other source); Gender = gender study (yes or no); Pop = type of population sampled (UG = undergraduate students; Par = parents; Oth = other); Rel = relationship
between participants (C/S = participant with own child and spouse; Dat = dating couple; Frn = friends; Oth = other; Res = researcher and participant; Spo = spouses or domestic partners; Str =
strangers); Comp = gender composition of group (Both = same gender and mixed gender combined; MG = mixed gender; SG = same gender; Oth = other); Size = group size (number of participants observed interacting together); Setting = observational setting (lab or other); Long = length of observation (in min); Res = whether the researcher was present during the observation
(no or yes); Rec = method of recording interaction (A = audiotape only; O = on-site coding only; V = videotape); Activity = type of activity observed (A-D = assigned decision making or debate
topic; A-T = assigned general topic(s); CH = interaction with child; DS = discussion about a disagreement; M-T = miscellaneous or mixed task activity; Mix = mixed activities; N-T = naturalistic task-oriented activity; SD = self-disclosure; Test = assigned test or puzzle; Unc = unclear; Unst = unstructured activity); N/A = information not available.
a. Carli (1989) tested for gender differences in affiliative speech using different operational definitions separately for same-gender groups, mixed-gender groups, and both groups combined.
The average effects associated with general affiliative speech for same-gender groups and for mixed-gender groups were used when using independent samples as the unit of analysis.
b. Reported separate tests of gender differences in affiliative speech using two or more different operational definitions with the same sample.
c. Reported findings from two separate studies with independent samples. For each study and sample, gender differences in affiliative speech were tested using two different operational
definitions.
d. Tested gender differences in affiliative speech using two operational definitions in two independent samples.
e. Reported separate tests of gender differences in affiliative speech in two or more independent samples.
f. McMullen et al. (1995) reported findings from two separate studies. The first study included two independent samples. The second study included one sample with two separate tests
using a similar operational definition (which was averaged when independent samples were used as the units of analysis).

Simkins-Bullock &
Wildman (1991)e
Stets & Burke (1996)b
Stets & Burke (1996)b
Stiles, Walz, Schroeder,
Williams, &
Ickes (1996)b
Stiles et al. (1996)b
Stiles et al. (1997)
Turner, Tjaden, &
Weismer (1995)
Turner et al. (1995)
W. Wood &
Karten (1986)
Yamada, Tjosvold,
& Draguns (1983)e
Yamada et al. (1983)e

Author and Year

TABLE 2:

341

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

64
30
38
16
32
16
32
16
32
16
32
32
32
32
32
5
5
60
24
24
84
84
84
79
79
29
54
36
10
10
25
22
33
76
25
25
24
24

64
30
30
16
32
16
32
16
32
16
32
32
32
32
32
5
5
60
24
24
84
84
84
79
79
29
54
36
10
10
25
16
17
76
19
19
24
24

Adams (1980)
Athenstaedt, Haas,
& Schwab (2004)c
Athenstaedt et al. (2004)c
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Carli (1989)a
Case (1988)b
Case (1988)b
Davis & Gilbert (1989)
Eskilson & Wiley (1976)
Fagot (1978)
Grotevant & Cooper
(1985)b
Grotevant & Cooper
(1985)b
Grotevant & Cooper
(1985)b
Hauser et al. (1987)b
Hauser et al. (1987)b
Heavey, Layne, &
Christensen (1993)
Heiss (1962)b
Heiss (1962)b
Hladik & Edwards
(1984)b
Hladik & Edwards
(1984)b
Jones, Gallois, Callan,
& Barker (1995)
Julien et al. (2000)c
Julien et al. (2000)c
Leaper (1987)
Leaper (1998)d
Leaper (1998)d
Leaper (1998)d
Leaper (1998)d

N (m)

N (w)

Dis
Crit
Crit
Inf
Sug
Dis
Sug
Dis

Dir

Inf

Dir
Inf
Crit

Sug
Sug
Inf

Inf

Inf

Dir
Dir
Task
Task
Task
Task
Dis
Dis
Dis
Dis
Task
Task
Dis
Dis
Dir
Crit
Sug
Task
Crit

Dir

Op Def

Study Characteristics for Assertive Language

Author and Year

TABLE 3:

0.19
0.19
0.21
0.37
0.08
0.39
0.33
0.08

F = 3.15
F = 3.04
F = 3.61
F = 10.67
F = 0.56
F = 4.37
p = .01
p = .28

0.32
0.71
0.52
0.15
0
0
0.13
0.14
0.29

M and SD
F = 6.22
M and SD
M and SD
p = .5
F = 0.02
F = 0.18
F = 0.49
F = 2.01

0.32

0.03
0.79
1.86
0.32
1.27
0
1.09
0.26
0.82
0.14
0
0.64
0
0
0.63
0.63
0.41
0.48
0.56

F = 0.02
F = 10.32
t = 5.27
t = 1.82
t = 3.58
t=0
t = 3.07
t = 1.46
t = 2.32
t = 0.81
p = .5
t = 2.57
p = .5
p = .5
p = .09
p = .09
F = 5.03
G
F = 7.61

M and SD

0.01

M and SD

Stat Value

W
W
W
M
M
M
M
M

M
M
M

M
M
M

W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W

Auth

Oth
Top
Top
Oth
Oth
Oth
Oth
Oth

Oth

Oth

Top
Oth
Oth

Top
Oth
Oth

Top

Top

Oth
Oth
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Oth
Oth
Top
Oth
Top

Top

Pub

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes

Yes

Yes
Yes
Yes

No
Yes
Yes

No

No

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes

Oth
Oth
Oth
UG
UG
UG
UG
UG

Par

Par

Oth
UG
UG

Par
Par
Par

Par

Par

UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
UG
Oth
Oth
UG
UG
Par

UG

Gender Pop

Str
Frn
Frn
Str
Frn
Frn
Frn
Frn

C/S

C/S

Spo
Dat
Dat

C/S
C/S
C/S

C/S

C/S

Res
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Str
Oth
Oth
Str
Str
C/S

Str

Rel

SG
SG
SG
MG
SG
SG
MG
MG

MG

MG

MG
MG
MG

MG
Both
Both

MG

MG

MG
SG
SG
MG
SG
MG
SG
MG
SG
MG
Both
Both
Both
Both
MG
MG
MG
Both
MG

Both

Comp

2
2
2
2
2
2
2
2

2
2
2

3
3
3

2
2
2
2
2
2
2
2
2
2
2
2
2
2
10
10
2
3
3

Size

Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab

Oth

Oth

Lab
Lab
Lab

Lab
Lab
Lab

Lab

Lab

Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Lab
Oth
Oth
Lab
Lab
Oth

Lab

8
20
20
5
10
10
10
10

30

30

14
20
20

20
N/A
N/A

20

20

20
20
20
20
20
20
20
20
20
20
20
20
20
20
45
45
7
N/A
300

N/A

Setting Long

No
No
No
No
No
No
No
No

No

No

No
No
No

No
No
No

No

No

No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Yes

No

Res

V
V
V
A
A
A
A
A

V
A
A

A
A
A

V
V
V
V
V
V
V
V
V
V
V
V
V
V
A
A
A
V
O

Rec

(Continued)

A-T
DS
DS
A-D
A-D
A-D
A-D
A-D

CH

CH

DS
A-D
A-D

A-D
A-D
A-D

A-D

A-D

A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
A-D
N-T
N-T
A-D
Test
C-N

A-D

Activity

342

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

32
9
9
39
39
39
39
35
35
35
35
20

12
24
96
96
96
96
18
18
18
76
76
76
76
20
20
57
21
18
18
20

32
9
9
39
39
39
39
37
37
37
37
20

12
24
72
72
72
72
18
18
18
76
76
76
76
20
20
57
21
18
18
20

Leaper et al. (1989)


Leik (1963)c
Leik (1963)c
Margolin & Wampold
(1981)b
Margolin & Wampold
(1981)b
Margolin & Wampold
(1981)b
Margolin & Wampold
(1981)b
Mikolic, Parker, &
Pruitt (1997)b
Mikolic et al. (1997)b
Mikolic et al. (1997)b
Mikolic et al. (1997)b
Moore, Shaffer, Goodsell,
& Baringoldz (1983)
Mulac, Wiemann,
Widenmann, &
Gibson (1988)d
Mulac et al. (1988)d
Nemeth, Endicott, &
Wachtler (1976)b
Nemeth et al. (1976)b
Nemeth et al. (1976)b
Nemeth et al. (1976)b
Pellegrini, Brody, &
Stoneman (1987)c
Pellegrini et al. (1987)c
Pellegrini et al. (1987)c
Piliavin & Martin (1978)b
Piliavin & Martin (1978)b
Piliavin & Martin (1978)b
Piliavin & Martin (1978)b
Pillon, Degauquier,
& Duquesne (1992)b
Pillon et al. (1992)b
Porter et al. (1985)
Reid, Keerie, &
Palomares (2003)
Resick et al. (1981)
Resick et al. (1981)
Robey, Canary, &
Burggraf (1998)

N (m)

N (w)

(Continued)

Author and Year

TABLE 3:

Crit

Oth
Dis
Crit

Inf
Dis
Sug

Inf
Inf
Inf
Task
Sug
Sug
Sug

Sug
Sug
Inf
Dis

Dir
Dir

Inf

Sug
Dir
Dis
Crit

Crit

Dis

Dir

Sug

Crit
Task
Task

Op Def

0.05
0.28

F = 0.09
F = 3.13

0.34
0.04
0.05
0.20

G
F = 0.03
F = 0.04
t = 0.9

0
0
0.60

0.36
0.38
0.46
0.03

p = .5
p = .5
F = 10.3

5.26
5.81
8.73
0.05
0.25
0.60
0.30
0.58
0.36
0.46
0.26

=
=
=
=

0.46
0.11

M and SD
M and SD
M and SD
F = 25.17
F = 9.77
F = 15.91
F = 5.22

F
F
F
F

M and SD
M and SD

p = .5

0.72
0.62
1.03
0.73

0.13

F = 0.62

9.27
6.95
18.99
9.52

0.19

F = 1.35

=
=
=
=

0
0.12
0.04

p = .5
PR
PR

F
F
F
F

Stat Value

M
W
W

W
W
W

M
M
M
W
W
W
W

U
U
U
U

M
M

M
M
M
M

M
M
M

Auth

Oth

Oth
Oth
Oth

Oth
Oth
Top

Oth
Oth
Oth
Oth
Oth
Oth
Oth

Oth
Oth
Oth
Oth

Oth
Oth

Oth

Top
Top
Top
Top

Top

Top

Top

Top

Oth
Oth
Oth

Pub

Yes

Yes
No
No

Yes
Yes
Yes

No
No
No
Yes
Yes
Yes
Yes

Yes
Yes
Yes
Yes

Yes
Yes

Yes

Yes
Yes
Yes
Yes

Yes

Yes

Yes

Yes

Yes
Yes
Yes

Oth

UG
Oth
Oth

UG
UG
UG

Par
Par
Par
UG
UG
UG
UG

UG
UG
UG
UG

UG
UG

UG

UG
UG
UG
UG

Oth

Oth

Oth

Oth

Par
Par
Par

Gender Pop

Spo

Str
Spo
Spo

Str
Str
Str

C/S
C/S
C/S
Str
Str
Str
Str

Str
Str
Str
Str

Str
Str

Str

Str
Str
Str
Str

Spo

Spo

Spo

Spo

C/S
Str
C/S

Rel

MG

MG
MG
MG

MG
MG
MG

MG
MG
MG
Both
Both
Both
Both

MG
MG
MG
MG

SG
MG

Both

Both
Both
Both
Both

MG

MG

MG

MG

MG
MG
MG

Comp

2
2
2

2
2
2

3
3
3
4
4
4
4

6
6
6
6

2
2

3
3
3
3

3
3
3

Size

Oth

Lab
Lab
Lab

Lab
Lab
Lab

Lab
Lab
Lab
Lab
Lab
Lab
Lab

Lab
Lab
Lab
Lab

Lab
Lab

Lab

Lab
Lab
Lab
Lab

Lab

Lab

Lab

Lab

Lab
Lab
Lab

10
7
7

10
10
10

36
36
36
35
35
35
35

120
120
120
120

20
20

N/A

50
50
50
50

20

20

20

20

N/A
N/A
N/A

Setting Long

No

No
No
No

No
No
No

No
No
No
No
No
No
No

No
No
No
No

No
No

No

Yes
Yes
Yes
Yes

No

No

No

No

No
No
No

Res

V
V
V

A
A
A

A
A
A
A
A
A
A

V
V
V
V

A
A

A
A
A
A

A
O
O

Rec

Unst

A-D
DS
DS

Mix
Mix
A-D

CH
CH
CH
A-T
A-T
A-T
A-T

A-D
A-D
A-D
A-D

A-D
A-D

Test

Test
Test
Test
Test

DS

DS

DS

DS

A-D
A-D
A-D

Activity

343

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

37
37
60
62
13
13
278
38
38
10
20
16
28
32

37
37
60
80
13
13
278
38
38
10
20
16
28
32

Task

Task
Task

Inf
Inf

Sug
Crit
Inf
Dir

Sug

Oth

Crit

Inf

Sug

0
0
0.14
0
0

P = 0.5
p = .5
p = .01
P = 0.5
p = .5

0.81
0
0.31

G
P = 0.5
F = 3.08

0.77
0.12

0.21

F = 1.51

M and SD
M and SD

0.1

0.39

F = 5.6
M and SD

0.34

F = 4.19

W
W

W
W

W
W
M
M

Oth

Top
Oth

Oth
Oth

Oth
Oth
Oth
Oth

Oth

Oth

Top

Oth

Oth

Yes

Yes
Yes

Yes
Yes

Yes
Yes
Yes
Yes

Yes

Yes

Yes

No

No

UG

UG
UG

UG
UG

UG
Oth
UG
UG

UG

UG

Oth

Par

Par

Str

Str
Str

Str
Str

Str
Spo
Str
Str

Str

Oth

Spo

C/S

C/S

MG

MG
SG

SG
MG

MG
MG
MG
MG

SG

SG

MG

MG

MG

4
2

2
2

2
2
2
2

Lab

Lab
Lab

Lab
Lab

Lab
Lab
Lab
Lab

Lab

Lab

Lab

Lab

Lab

N/A

5
N/A

5
5

15
15
6
6

15

N/A

N/A

No

No
No

No
No

No
No
No
No

No

No

No

No

No

V
V

V
V

A
V
V
V

Unc

A-D
Unc

A-D
A-D

A-D
DS
Unst
Unst

A-D

A-D

DS

CH

CH

NOTE: N (w) = number of women; N (m) = number of men; Op Def = operational definition (Crit = criticize; Dir = directs or commands; Dis = disagrees or challenges; Inf = gives information; Sug = suggestion; Task = general task-oriented statements; Oth = other) Stat value = statistical value (M and SD = effect size computed from reported means and standard deviations; G = effect size computed based on aggregate of similar measures for same operational definition); G = index of effect size; Auth = first authors gender (M = man; W = woman; U =
unclear); Pub = publication source (Top = top journal; Oth = other source); Gender = gender study (yes or no); Pop = type of population sampled (UG = undergraduate students; Par =
parents; Oth = other); Rel = relationship between participants (C/S = participant with own child and spouse; Dat = dating couple; Frn = friends; Oth = other; Res = researcher and participant; Spo = spouses or domestic partners; Str = strangers); Comp = gender composition of group (Both = same gender and mixed gender combined; MG = mixed gender; SG = same gender
and mixed gender combined); Size = group size (number of participants observed interacting together); Setting = observational setting (lab or other); Long = length of observation (in min);
Res = whether the researcher was present during the observation (no or yes); Rec = method of recording interaction (A = audiotape only; O = on-site coding only; V = videotape); Activity =
type of activity observed (A-D = assigned decision making or debate topic; A-T = assigned general topic(s); CH = interaction with child; DS = discussion about a disagreement; M-T = miscellaneous or mixed task activity; Mix = mixed activities; N-T = naturalistic task-oriented activity; SD = self-disclosure; Test = assigned test or puzzle; Unc = unclear; Unst = unstructured
activity); N/A = information not available.
a. Carli (1989) tested for gender differences in assertive speech using different operational definitions separately for same gender groups, mixed gender groups, and both groups combined.
The average effects associated with general assertive speech for same-gender groups and for mixed-gender groups were used when using independent samples as the unit of analysis.
b. Reported separate tests of gender differences in assertive speech using two or more different operational definitions with the same sample.
c. Reported separate tests of gender differences in assertive speech in two or more independent samples.
d. Tested gender differences in assertive speech using two or more operational definitions in two independent samples.
e. McMullen et al. (1995) reported findings from two separate studies with independent samples. For each studys sample, there were two separate tests using a similar operational
definition (which were averaged when independent samples were used as the units of analysis).

Roopnarine & Adams


(1987)b
Roopnarine & Adams
(1987)b
Sayers & Baucom
(1991)
Scudder & Andrews
(1995)
Simkins-Bullock &
Wildman (1991)c
Simkins-Bullock &
Wildman (1991)c
Stets & Burke (1996)
Stiles (1996)b
Stiles (1996)b
Turner, Tjaden, &
Weismer (1995)
Turner et al. (1995)
W. Wood & Karten
(1986)
Yamada et al. (1983)c
Yamada, Tjosvold, &
Draguns (1983)c

344

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

often published in other sources, these top-tier journals


consistently publish work considered of very high quality, and this classification method did not require
making subjective evaluations of a given studys quality.
Statistical Analyses

heterogeneous level of the original variable. This is analogous to testing for simple main effects in an ANOVA.
When significant heterogeneity remained at particular
levels of a moderator, follow-up tests were performed.
That is, we explored if the effect of one moderator was
specific to the particular condition of a second moderator. However, none of these analyses revealed any noteworthy findings, and therefore are not described later.

Effect Sizes
B. Johnsons (1989, 1993) DSTAT software was
used to carry out the statistical analyses. DSTAT reports
Cohens d index of effect size, which represents the
average group difference in standard deviation units.
According to Cohen (1988), an effect size below .2 (i.e.,
more than 85% overlap between women and men) is
considered negligible. Meaningful effect sizes are characterized as small when d .2 (i.e., 85% or less overlap
between women and men), medium when d .5 (i.e.,
66% or less overlap), and large when d .8 (i.e., less
than 53% overlap). Based on the hypothesized direction
of gender differences, average effects were positive (1) if
women were higher than men in talkativeness, (2) if
women were higher than men in affiliative speech, and
(3) if men were higher than women in assertive speech.

Inferential Statistics
The DSTAT software converts the inferential statistics used to test for a gender difference (e.g., t, F, r, p, or
M and SD) into Hedges g standardized effect measure.
DSTAT subsequently computes the combined effect size
(Cohens d or correlation coefficient r) across studies as
well as focused comparison tests of effect sizes on
blocked moderator variables using fixed effects.
Between-study variance of blocked variables is modeled
using the analog to the analysis of variance technique
(B. Johnson, 1989). Similar to the one-way ANOVA,
this technique handles categorical independent variables
that are used to group effect sizes into mutually exclusive categories (Lipsey & Wilson, 2001). This procedure
partitions the total variance into the portion explained
by the categorical variable (QB), and the residual pooled
within-groups portion (QW). Each of these Q values is
distributed as a chi-square statistic. Interpretation of the
fit of the categorical models is twofold: (1) a significant
QB statistic indicates that the mean effect sizes across
groups differ by more than sampling error; (2) a nonsignificant QW indicates homogeneity of effect sizes
within the groups indicating no further variation among
effect sizes. If a model indicates a significant QW for a
level of a categorical variable, that variable alone is not
sufficient for understanding its constituent effect sizes
(Lipsey & Wilson, 2001). Thus, the influence of other
independent variables may be investigated within the

Trimming
To ensure that the overall effect sizes for each of the
analyses accurately represented the overall distribution
of effects, a trimming procedure was used to examine
the stability of the overall effect size. Two separate
analyses were performed to exclude the most extreme
10% and 20% of sampled studies to test overall effects
with these reduced samples. These procedures did not
appreciably alter the findings; thus, overall effect sizes
were not dependent on a small proportion of the
samples included in the meta-analyses (see Results).

Units of Analysis
Independent samples. The independent sample is a
unit of analysis based on each independent group for
which a gender comparison was made. For example,
several studies reported gender effects separately for
same-gender and mixed-gender groups. These were
treated as two independent samples and entered separately into the meta-analysis. The independent sample
was the unit of analysis used to test each of the moderator variables except for operational definition
(explained next).
Test. In some studies, more than one measure of a
given language construct was analyzed. For example, a
study may have tested different types of assertive speech.
When test is used as the unit of analysis, the effect sizes
for each individual statistical test for a given language
construct are included. Studies with more than one operational definition of the construct attain more weight in
the average computation of the effect than those that
include only one operational definition. Therefore, test is
not an appropriate unit of analysis when analyzing other
moderators. This unit of analysis was used only when
examining operational definition as a moderator. In the
other analyses, the effect sizes were averaged when there
was more than one test for a sample.

File Drawer Problem


Some authors advocate including unpublished studies in meta-analyses out of concern for the file-drawer
problem. This refers to a potential bias toward

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


TABLE 4:

Stem and Leaf Display of Mean Effect Sizes for Gender


Differences in Talkativeness

Stem
2.1
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

TABLE 5:

345

Stem and Leaf Display of Mean Effect Sizes for Gender


Differences in Affiliative Talk

Stem
4

0
0
0
1
4
6
0
7
2
3
0
0
8
3
9
2
9
0
7

1
4
8
1
8

67
6
224
99

4556
77
88
00000000000000000009
67
5689

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
TABLE 6:

06
3
4
2
128
4
0
3
9
1
2
9
7
0

6678
00000000000000077
559
1379
478

2448

6
Stem and Leaf Display of Mean Effect Sizes for Gender
Differences in Assertive Talk

Stem

the publication of significant results with many null


results going unpublished (and thereby residing in
researchers file drawers). Tracking down unpublished
studies, however, can be time-consuming and expensive
(in cases of ordering microfiches). Some researchers
have called into question the necessity of including
unpublished studies in all meta-analyses. Recent reviews
indicate that it is common to find meta-analyses based
only on published studies (Sharpe, 1997), and that comparisons of meta-analyses with published and unpublished studies usually do not differ in results (Sutton,
Duval, Tweedie, Abrams, & Jones, 2000).

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2

7
0
6
1
6
5
0
0
0
1
0
0
1
2
0
0
1

2
9
2
1
0
2
1
1
6

4
7
0
2
3
4
8

6
000000347
5
6

3
19

There are four reasons why the file-drawer problem


does not appear relevant in the present set of meta-analyses. First, including unpublished studies is primarily
warranted when reviewing a research area with few
published studies (Sharpe, 1997)which was not
the case here (see Results). Second, the present set of

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

346

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

TABLE 7:

Gender Effects on Talkativeness, Affiliative Speech, and Assertive Speech for Independent Samples

Talkativeness
Affiliative speech
Assertive speech

95% Confidence Interval

QW

70
54
50

4,385
2,781
2,541

.14***
.12***
.09***

.19/.08
.06/.18
.02/.15

.07
.06
.04

243.42***
99.27***
83.47***

NOTE: Positive effect sizes reflect higher scores for women with talkativeness, women with affiliative speech, and men with assertive speech.
Thus, the negative average effect size associated with talkativeness indicates that contrary to prediction, men tended to be more talkative than
women.
***p < .01.

sampled studies indicated no shortage of null results.


This is depicted in the stem-and-leaf plots of the effect
sizes for each of the meta-analyses presented in Tables
4, 5, and 6. This point is underscored by the negligible
average effect sizes for all three language measures (see
Results). Finally, if there was a publication bias toward
reporting significant gender effects, then one might
expect the bias to be more likely in studies focusing on
gender than those that did not. This was not found with
any of the three language constructs (see Results). Thus,
the analyses do not indicate that there was a file-drawer
problemwith the prevalence of null results perhaps
being the most compelling point.

effect size indicates that contrary to prediction, men were


significantly more talkative than were women.

Trimming
When 10% of the sampled scores were trimmed,
there were 63 remaining samples with a statistically significant average effect size of d = .11 (95% CI = .17 /
.05). When 20% of sampled scores were trimmed,
there were 56 remaining studies with a statistically significant gender difference and an effect size of d = .14
(95% CI = .20 / .08). Thus, trimming 10% or 20% of
the scores did not appreciably affect the overall finding,
which means that there was no apparent bias from
outlier scores.

RESULTS

Moderators
Three sets of meta-analyses were performed examining average gender differences in talkativeness, affiliative language, and assertive language, respectively.
Furthermore, focused comparison tests of significance
levels and effect sizes were carried out for various moderators with each language variable. The results are
summarized in Tables 7 to 11. Table 7 presents the
average gender effects on talkativeness, affiliative
speech, and assertive speech. Table 8 presents the results
associated with the operational definition moderator
for each language construct. The effects of the other categorical moderator variables are presented separately
for talkativeness, affiliative speech, and assertive speech
in Tables 9, 10, and 11, respectively.
Talkativeness

Overall Gender Effects


The meta-analysis of gender differences in talkativeness was based on a total of 70 independent samples with
a total N = 4,385 participants. We tested the hypothesis
that women would be more talkative than would men.
Among these investigations, there was a statistically significant mean effect size of d = .14 (95% confidence
interval [CI] = .19 / .08). The negative direction of the

When classifying the relationship between partners,


some studies combined familiar partners and strangers,
which we refer to here as mixed-familiarity groups.
Some independent samples were excluded when testing
certain moderators as follows (with the reasons in
parentheses): author gender, k = 1 (only initials for first
name); length of observation, k = 10 (9 studies with
missing information and 1 outlier with 2,880 min);
researchers presence, k = 1 (unclear procedure); setting,
k = 1 (unclear setting); relationship, k = 2 (1 was a study
between professor and students; 1 did not describe the
nature of the relationship); group size, k = 1 (size not
specified); gender composition, k = 23 (combined scores
for same and mixed gender); activity, k = 5 (mixed
activities or unclear activity).
The analyses revealed seven significant (or marginally significant) moderators of gender differences in
talkativeness speech (see Tables 8 and 9). First, operational definition accounted for variations in effect sizes.
The effect sizes associated with measures using MLU
(d = .37), total statements (d = .28), and duration
(d = .24) were significantly different than were measures using either total words (d = .01) or total turns
(d = .40). Rate (d = .14) was significantly different
from total turns but not any of the other measures. The

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


TABLE 8:

347

Gender Effects on Talkativeness, Affiliative Speech, and Assertive Speech by Operational Definition

Type of Analysis
Talkativeness
Total turns
Total words
Rate
Duration
Miscellaneous
Total statements
Mean length of utterance
Affiliative speech
Active understanding
General socioemotional
Support
Agree
Acknowledge
Assertive speech
Criticize
Disagreements
Directives
Informs
Miscellaneous
Suggestions
General task

95% Confidence Interval

QW

4
18
9
33
1
4
8

239
1,350
571
2,107
20
220
313

.40a**
.01b
.14bc
.24c**
.26abc
.28c**
.37c**

.15/.65
.09/.11
.29/.01
.32/.16
.88/.36
.53/.03
.56/.17

.00
.00
.07
.12
.13
.14
.18

5.98
44.44***
27.17***
100.11***
0.00
0.86
37.53***

3
16
10
13
26

152
667
554
818
1,577

.41a**
.35a**
.16ab**
.08b
.01b

.10/.72
.21/.48
.01/.32
.03/.18
.09/.08

.20
.17
.08
.04
.00

4.07
29.43**
26.15***
33.18***
30.95

12
10
11
15
2
13
9

672
499
467
779
163
934
386

.13a**
.06ab
.00ab
.07b
.23bc
.27c**
.38c**

.25/.02
.22/.09
.15/.16
.04/.19
.06/.53
.16/.38
.20/.56

.07
.03
.00
.04
.12
.13
.19

15.07
29.64***
20.51**
21.32
0.12
20.28*
18.26**

QB
38.45***

24.47***

40.32***

NOTE: The moderating effect of operational definition on gender differences in each of the language constructs was analyzed using test as the unit
of analysis. With talkativeness and affiliative speech, a positive effect size indicates a higher mean score for women than men. With assertive speech,
a positive effect size indicates a higher mean score for men than women.
*p .10. **p < .05. ***p < .01.

operational definitions associated with significant effect


sizes included MLU, total statements, duration, and
total turns. Of these, total number of turns was the only
measure associated with greater talkativeness among
women than men. The latter effect is based on only four
samples and therefore should be viewed cautiously.4
Second, the relationship between the participants
was significant. A sizable positive effect size (i.e.,
women more talkative) was seen in studies of classmates
(d = .54) and parents with their children (d = .42),
which were significantly different than the negative
effect sizes (i.e., men more talkative) seen between
spouses or partners (d = .38), mixed-familiarity groups
(d = .23), and strangers (d = .17). Also, interactions
between friends (d = .06) were significantly different
than those between classmates or spouses/partners.
(Only one study examined dating partners, and therefore it is not mentioned here.) A follow-up analysis was
performed to test the overall effect of familiarity.
Friends, dating partners, spouses or partners, and
parents with children were combined to create a closerelationship category. In the subsequent test, interactions between close relationships (d = .04) were
different than between strangers, QB(1) = 5.07, p < .05;
classmates, QB(1) = 6.11, p < .05; or mixed-familiarity
groups, QB(1) = 3.79, p = .05.

Third, group size was a marginally significant (p <


.10) moderator of gender differences in talkativeness.
Mens greater talkativeness was more likely during
dyadic interactions (d = .16) than group interactions (d
= .07). Fourth, effect sizes were significantly larger in
mixed gender groups (d = .28) than same gender
groups (d = .08). Fifth, gender effects were larger with
studies that took place in a university lab (d = .17) than
those that took place elsewhere (d = .03).
Sixth, effect sizes varied with the activity. There
were small to moderate and positive effect sizes during
child-oriented activities (d = .43) or self-disclosure
(d = .32). Both of these activities reflect femininestereotyped socioemotional contexts. Given this commonality as well as the small number of studies in each
activity, we combined the two activities (d = .39) to
make subsequent contrasts. They were significantly
different than the moderate and negative effect sizes
associated with discussions of nonpersonal topics (d =
.79), QB(1) = 32.74, p < .01; or disagreements (d =
.56), QB(1) = 14.25, p < .01. Child-oriented, self-disclosure, nonpersonal topics, and disagreements were
all significantly different from the negligible effect
sizes associated with miscellaneous tasks (d = .05),
unstructured activities (d = .07), or deliberations
(d = .13).

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

348

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

TABLE 9:

Comparison Tests of Moderator Variables of Gender Effects on Talkativeness

Moderator
INTERACTIVE CONTEXT
Student status
Undergraduate
Other
Relationship
Classmates
Parents and child
Dating partners
Friends
Strangers
Mixed familiarity
Spouses/partners
Group size
Dyad
Group
Gender composition
Same gender
Mixed gender
Researcher present
Yes
No
Setting
Lab
Other
Activity
Child oriented
Self-disclosure
Miscellaneous tasks
Unstructured
Deliberation
Disagreement
Nonpersonal topic

54
16

3,011
1,374

3
2
1
3
50
5
4

95% Confidence Interval

QW

.15**
.10a

.22/.09
.20/.00

.08
.05

197.62***
45.16***

111
50
54
174
3,303
302
92

.54a**
.42ab**
.32ab
.06bc
.17cdc**
.23cd**
.38d**

.17/.91
.03/.82
.06/.70
.22/.34
.23/.10
.43/.02
.67/.08

.26
.21
.16
.03
.08
.11
.19

2.45
0.70
0.00
4.38
187.48***
4.96
9.48*

49
19

2,783
1,261

.16a**
.07a

.23/.10
.18/.04

.08
.03

179.12***
59.85***

13
34

669
1,630

.08a
.28b**

.22/.09
.36/.20

.04
.14

71.17***
94.64***

9
60

623
3,658

.24a**
.11a**

.40/.07
.17/.05

.12
.06

56.45***
180.35***

58
11

3,220
1,160

.17a**
.03b

.23/.10
.15/.08

.08
.02

227.24***
11.37

2
2
11
11
30
2
7

50
60
1,291
652
1,757
57
199

.43a**
.32ab
.05b
.07b
.13b**
.56c**
.79c**

.03/.83
.20/.83
.16/.06
.22/.08
.22/.05
.94/.18
1.05/.54

.21
.16
.03
.03
.07
.27
.37

0.70
2.39
3.34
38.77***
114.81***
5.68*
26.71***

2,197
2,035
153

.16a**
.12a**
.11a

.24/.08
.20/.04
.37/.16

.08
.06
.05

82.83***
141.60***
18.53***

1,500
1,258
876

.19a**
.05a
.08a

.28/.10
.16/.05
.21/.04

.09
.03
.04

42.10***
81.31***
90.93***

2,929
1,448

.13a**
.18a**

.20/.06
.25/.06

.06
.08

176.10***
67.05***

3,893
492

.13a**
.18a**

.19/.07
.36/.01

.07
.09

222.26***
20.84***

859
3,526

.36a**
.08b**

.48/.24
.14/.02

.18
.04

48.08***
179.36***

2,638
1,747

.14a**
.14a**

.21/.06
.22/.05

.07
.07

139.05***
104.37***

QB
0.65

32.97***

3.15*

5.39**

1.87

4.22**

44.38***

METHODOLOGICAL FACTORS
Type of recording
Video
35
Audio
32
On site
3
Length
3-9 min
21
10-18 min
20
20-120 min
19
First-author gender
Woman
42
Man
27
Gender study
Yes
62
No
8
Publication source
APA journal
13
Other source
57
Publication year
1960-1985
35
1986-2005
35

0.46

4.12

0.28

0.32

15.97***

0.04

NOTE: APA = American Psychological Association. For each moderator variable, mean effect sizes (d) with different subscripts are significantly
different (p < .05).
*p .10. **p < .05. ***p < .01.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


TABLE 10:

349

Comparison Tests of Moderator Variables of Gender Effects on Affiliative Speech

Moderator
INTERACTIVE CONTEXT
Student status
Undergraduate
Other
Relationship
Classmates
Dating partners
Strangers
Friends
Parents and child
Mixed familiarity
Spouses/partners
Group size
2
3 or more
Gender composition
Same gender
Mixed gender
Researcher present
Yes
No
Setting
Lab
Other
Activity
Nonpersonal topic
Deliberation
Self-disclosure
Miscellaneous tasks
Child oriented
Disagreement
Unstructured

95% Confidence Interval

QW

37
17

1,800
981

.20a**
.01b

.12/.29
.11/.08

.10
.01

69.91***
6.64

2
1
32
6
6
1
6

103
54
1,620
210
321
44
429

.34ab
.32ab
.18a**
.09ab
.07ab
.00ab
.06b

.05/.73
.06/.70
.09/.27
.15/.34
.11/.24
.59/.59
.20/.08

.17
.16
.09
.05
.03
.00
.03

1.92
0.00
64.46***
5.07
0.93
0.00
3.05

40
14

1,889
892

.09a**
.17a**

.01/.16
.05/.29

.04
.08

54.25
31.76***

12
27

510
1,062

.33a**
.01b

.15/.51
.07/.10

.16
.01

20.91**
33.50

52
1

2,637
24

.11a**
.00a

.04/.17
.57/.57

.05
.00

85.80***
0.00

50
4

2,720
61

.12a**
.18a

.06/.19
.54/.18

.06
.09

82.81***
1.78

2
23
2
6
2
5
9

202
999
54
402
61
465
344

.44a**
.21bc**
.20abc
.03c
.00c
.03c
.10c

.16/.72
.10/.32
.26/.67
.16/.23
.35/.35
.16/.11
.28/.09

.21
.10
.10
.02
.00
.01
.05

2.53
43.29***
1.05
10.14
0.00
1.36
4.58

1,025
1,675
81

.16a**
.08a
.11a

.05/.26
.00/.17
.19/.42

.08
.04
.06

24.14
61.63***
0.38

538
834
867

.21a**
.01b
.18a**

.07/.35
.11/.12
.07/.30

.10
.00
.10

22.87*
6.64
30.17***

1,703
1,078

.14a**
.07a

.06/.22
.04/.17

.07
.03

53.80***
32.12*

2,588
193

.13a**
.02a

.06/.19
.24/.20

.06
.01

85.71***
0.09

519
2,262

.14a**
.10a**

.00/.28
.03/.18

.07
.05

21.13**
65.91***

1,099
1,682

.20a**
.06b

.10/.31
.03/.14

.10
.03

38.02***
44.45

QB
10.66***

11.79*

1.20

10.09***

0.13

2.63*

18.42***

METHODOLOGICAL FACTORS
Type of recording
Audio
21
Video
29
On site
4
Observation length
4-8 min
14
10-15 min
13
20-300 min
16
First-author gender
Woman
32
Man
22
Gender study
Yes
51
No
3
Publication source
APA journal
11
Other source
43
Publication year
1960-1985
20
1986-2005
34

1.08

6.54**

1.30

1.42

0.19

4.76**

NOTE: APA = American Psychological Association. For each moderator variable, mean effect sizes (d) with different subscripts are significantly
different (p < .05).
*p .10. **p < .05. ***p < .01.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

350

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

TABLE 11:

Comparison Tests of Moderator Variables of Gender Effects on Assertive Speech

Moderator
INTERACTIVE CONTEXT
Student status
Undergraduates
Other
Relationship
Classmates
Strangers
Dating partners
Friends
Parents and child
Spouses/partners
Group size
2
3 or more
Gender composition
Same gender
Mixed gender
Researcher present
Yes
No
Setting
Lab
Other
Activity
Nonpersonal topic
Deliberation
Unstructured
Miscellaneous tasks
Disagreement
Child oriented

29
21

1,537
1,004

2
26
1
5
10
6

95% Confidence Interval

QW

.19a
.03b

.10/.27
.13/.06

.09
.02

44.64*
27.75

147
1,302
54
186
408
444

.23a
.22a**
.12abc**
.05abc
.00bc
.12c

.10/.55
.12/.31
.26/.50
.21/.31
.15/.15
.25/.01

.11
.11
.06
.03
.00
.06

0.20
45.85***
0.00
2.61
13.42
2.64

33
17

1,663
878

.02a
.11a**

.00/.15
.01/.23

.04
.06

53.59**
29.62**

11
33

550
1,393

.29a**
.03b

.12/.46
.05/.11

.15
.01

21.70**
42.86

2
48

96
2,445

.46a**
.10b**

.83/.10
.04/.17

.23
.05

0.13
74.26**

46
4

2,482
59

.10a**
.24a

.03/.16
.61/.12

.05
.12

77.55***
2.71

2
26
2
4
7
6

202
1,371
58
165
512
125

.48a**
.20b**
.07b
.01b
.10b*
.28b**

.20/.76
.11/.29
.43/.30
.32/.29
.22/.03
.53/.03

.23
.10
.03
.01
.05
.14

0.71
34.18
0.24
6.19
5.97
4.95

1,319
1,141
81

.14a**
.06ab
.22b

.04/.23
.03/.16
.53/.09

.07
.03
.11

27.24
49.08***
2.09

355
512
860

.08a
.02a
.13a**

.08/.24
.15/.10
.02/.24

.04
.01
.07

19.18**
15.58*
40.96***

1,342
1,031

.11a**
.04a

.02/.20
.06/.14

.05
.02

61.88***
19.17

2,348
193

.10a**
.03a

.03/.17
.23/.17

.05
.01

74.15***
7.92

QB
11.08***

18.88***

0.26

7.64**

9.08***

3.21*

30.30***

METHODOLOGICAL FACTORS
Type of recording
Audio
25
Video
21
On site
4
Observation length
5-8 min
10
10-15 min
9
20-300 min
20
First-author gender
Woman
27
Man
22
Gender study
Yes
44
No
6
Publication source
APA journal
14
Other source
36
Publication year
1960-1985
16
1986-2005
34

5.07*

3.29

1.12

1.40

0.30
721
1,820

.11a
.05a

.01/.23
.00/.15

.06
.04

39.22***
43.95

928
1,613

.18a**
.04b

.07/.29
.04/.12

.09
.02

18.98
60.60***

3.89**

NOTE: APA = American Psychological Association. For each moderator variable, mean effect sizes (d) with different subscripts are significantly different (p < .05).
*p .10. **p < .05. ***p < .01.

Finally, besides operational definition (previously


described), publication source was another significant
methodological moderator. Studies published in top-tier
journals (d = .36) had larger effect sizes than those
published in other sources (d = .08).

Follow-Up Regression Analysis


To explore the relative impact of these factors, multiple regression was carried out using the standardized
g measures of effect size. Contextual factors that

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


significantly moderated gender differences in talkativeness (described above) were dummy coded for the
regression as follows: familiarity (1 = stranger, 0 =
other), gender composition (1 = mixed gender, 0 =
other), setting (1 = lab, 0 = other), and activity (1 = nonpersonal topic or disagreement, 0 = other). The resulting model was significant, R2 = .23, F(4, 65) = 4.91, p <
.01. Two significant predictors occurred in the model,
which were activity, = .37, t = 3.33, p < .01; and
gender composition, = .25, t = 2.14, p < .05. (As previously explained, the average effect size for gender differences in talkativeness was negative, with men generally
being more talkative; therefore, the betas indicate that these
two factors were related to mens greater talkativeness.)
Affiliative Speech

Overall Gender Effects


Across all studies testing for gender differences in affiliative speech, there was a total of 54 independent samples
with a total N = 2,781. The hypothesis was that women
would use more affiliative speech than would men. Among
these investigations, there was a significant mean effect size
of d = .12 (95% CI = .06 / .18). The positive direction of
the effect size indicates that as predicted, women used significantly more affiliative speech than men.

Trimming
When 10% of the sampled scores were trimmed, there
were 49 remaining studies with a statistically significant
gender difference with an effect size of d = .08 (95%
CI = .01 / .15). When 20% of sampled scores were
trimmed, there were 44 remaining studies with a nonsignificant effect size of d = .03 (95% CI = .04 / .10).
Thus, trimming 10% or 20% of the scores did not appreciably affect the magnitude of the average effect size.

351

socioemotional speech were significantly larger than


those for agreement (d = .08) and acknowledgment
(d = .01). The effect size for supportive speech did not
significantly differ from any other definitions.
The results implicated five contextual moderators of
gender differences in affiliative speech. First, gender differences were more likely in studies of undergraduates
(d = .20) than those focusing on other populations (d =
.01). Second, gender differences among strangers (d = .18)
was significantly greater than the difference among
spouses or partners (d = .06). More generally, when close
relations were combined into a single category (friends,
dating partners, spouses, spouse and children), the contrast between strangers (d = .20) and close relations
(d = .02) was significant, QB(1) = 5.78, p < .05. Third,
there was a significant gender difference among studies of
same-gender groups (d = .33), but not those of mixedgender groups (d = .01). Fourth, there was a trend (p =
.10) suggesting that studies in university labs (d = .12) differed from those in other settings (d = .18). Finally,
gender differences in affiliative speech were sizable among
studies looking at conversations about nonpersonal topics
(d = .44), self-disclosures (d = .20), and deliberations (d =
.21). These were significantly different than the negligible
differences associated with studies examining miscellaneous tasks (d = .03), child-oriented activities (d = .00), disagreements (d = .03), or unstructured activities (d = .10).
Besides operational definition, publication year and
observation length were two other methodological
moderators. First, effect sizes were significantly larger
among studies published prior to 1985 (d = .20) than
those published more recently (d = .06). Second, there
was a curvilinear association between observation
length and effect size. Gender differences occurred in
studies based on very brief (4-8 min) observations (d =
.21) and somewhat long (20-300 min) observations (d =
.18), but not in studies based on somewhat short (10-15
min) observations (d = .01).

Moderators
We excluded some independent samples when testing
the following moderators for the cited reasons: gender
composition, k = 15 (combined scores for same and mixed
gender); researchers presence, k = 1 (unclear procedure);
length of observation, k = 11 (information not indicated);
and activity, k = 5 (mixed or unclear activities).
There were eight significant or marginally significant
moderators of gender differences in affiliative speech
(see Tables 8 and 10). First, the operational definition
predicted variations in effect sizes. Significant average
effect sizes (i.e., women greater than men) occurred
with general active understanding (d = .41), general
socioemotional speech (d = .35), and support (d = .16).
The effect sizes for active understanding and general

Follow-Up Regression Analysis


Once again, we performed a multiple regression to test
the relative influence of the significant contextual moderators. We dummy coded the following factors that were
previously indicated to be significant moderators of
gender differences in affiliative speech: student status (1 =
undergraduate, 0 = other), gender composition (1 = same
gender, 0 = other), and activity (1 = nonpersonal topic,
deliberation, or self-disclosure, 0 = other). The resulting
model was significant, R2 = .24, F(3, 50) = 5.17, p < .01.
Two of the predictors were significant: activity, = .27,
t = 2.12, p < .05; and gender composition, = .26,
t = 2.05, p < .05. In addition, student status was marginally
significant, = .23, t = 1.85, p < .08.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

352

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

Assertive Speech

Overall Gender Effects


Among those studies testing for gender differences in
assertive speech, there was a total of 50 independent
samples with a total N = 2,541. Assertive speech was
hypothesized to be more likely among men than
women. There was a statistically significant but negligible mean effect size of d = .09 (95% CI = .02 / .15). The
positive direction of the effect size indicates that as predicted, men used significantly more assertive speech
than did women.

Trimming
When 10% of the sampled scores were trimmed,
there were 45 remaining studies with a statistically significant gender difference with an effect size of d = .09
(95% CI = .02 /.17). When 20% of sampled scores were
trimmed, there were 40 remaining studies with a statistically significant gender difference and an effect size of
d = .13 (95% CI = .05 / .20). Thus, trimming 10% or
20% of the scores indicated no appreciable bias from
outlier scores.

Moderators
Some independent samples were excluded when testing certain moderators as follows: gender composition,
k = 6 (studies combined scores for same and mixed
gender); activity, k = 3 (unclear activity or mixed activities); observation length of, k = 11 (information not
indicated); and first-author gender, k = 1 (ambiguous
first name).
There were several significant or marginally significant
moderators of gender differences in assertive speech (see
Tables 8 and 11). First, operational definition was a significant factor. Effect sizes were significantly larger for
general task-oriented speech (d = .38) and suggestions
(d = .27) than other measures; also, both of these measures had significant average effect sizes. Criticism (d =
.13) also had a significant average effect size, and the
difference favored women. There were no average differences associated with informing (d = .07), directives
(d = .00), or disagreements (d = .06).
Besides operational definition, there were six significant contextual moderators and two other methodological moderators of gender differences in assertive speech.
Among the contextual factors, first, average gender differences were larger among undergraduates (d = .19) than
among others (d = .03). Second, interactions between
classmates (d = .23) or strangers (d = .22) were significantly different from those of parents with their children
(d = .00) or spouses/partners (d = .12). Also, when all
close relationships (friends, dating partners, spouses,

spouses with children) were combined, there was a


significant difference between strangers (d = .22) and
close relations (d = .04), QB(1) = 15.18, p < .01. Third,
gender differences were larger for same-gender groups
(d = .29) than for mixed-gender groups (d = .03). Fourth,
effect sizes were significantly different depending on
whether the researcher was present (d = .46) or not (d =
.10) during the interaction. (However, the researcher was
present in only 2 out of 50 studies, and therefore this
result should be viewed cautiously.) Fifth, studies occurring in a research lab (d = .10) tended to be different (p <
.10) than those occurring elsewhere (d = .24).
Finally, the activity was the last aspect of the interactive context that moderated gender differences in
assertive speech. There was a significant difference
between studies examining discussions of nonpersonal
topics (d = .48) than other activities. This difference
should be viewed cautiously, however, because only
two studies looked at nonpersonal topics. Perhaps more
meaningful is that significant and positive effect sizes in
assertive speech (i.e., men higher than women) occurred
only among studies examining discussions of nonpersonal topics (d = .48) or deliberations (d = .20). In contrast, significantly negative effect sizes (i.e., women
higher than men) occurred in studies of child-oriented
activities (d = .28). Negligible effect sizes were
associated with unstructured activities (d = .07) and
disagreements (d = .10).
In addition to operational definition (described earlier), one other methodological factor was implicated.
Effect sizes were significantly larger among older studies (d = .18) than more recent studies (d = .04).

Follow-Up Regression Analysis


A regression analysis tested the relative influence of the
significant contextual moderators of gender differences in
assertive speech. For this purpose, we dummy coded the
following factors: student status (1 = undergraduate, 0 =
other), familiarity (1 = stranger or classmates, 0 = close
relations), gender composition (1 = same gender, 0 =
other), and activity (1 = nonpersonal topic or deliberation,
0 = other). A significant regression model was obtained,
R2 = .24, F(4, 45) = 3.64, p < .05. Familiarity was a significant predictor, = .41, t = 2.22, p < .05; and activity
was marginally significant, = .27, t = 1.85, p < .08.
DISCUSSION
Statistically significant average gender differences in
talkativeness, affiliative speech, and assertive speech
were revealed in the present meta-analyses. However,
the most interesting findings were the influence of the
various moderators on these gender differences. We will

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


first review the overall gender differences in language
use. Then we will focus on the impact of the various
moderator variables. We will also consider the relevance
of the results to possible explanations for gender differences in language use.
Overall Gender Differences
Separate sets of meta-analyses were carried out to
examine possible gender differences in talkativeness,
affiliative speech, and assertive speech. Average effect
sizes were statistically significant, though negligible (d <
.2), for each of the three measures. Contrary to prediction, men tended to talk significantly more than women.
However, as predicted, women tended to use more affiliative speech. Also, a negligible effect size was found for
assertive speech, with men statistically more likely than
women to use assertive speech. As discussed later, more
meaningful effect sizes occurred when particular factors
were taken into account. But given the small average
effect sizes, we stress that womens and mens verbal
behavior were generally quite similar.
The finding that men tended to be more talkative than
women contradicts the popular stereotype that women
are more verbose than men (see James & Drakich, 1993)
as well as a meta-analysis suggesting that females are
more verbally skilled than males (Hyde & Linn, 1988).
The observed pattern is also opposite to the finding from
a meta-analysis comparing mothers and fathers talkativeness with their children. Leaper et al. (1998) found an
average effect size of d = .20 with mothers being more
talkative than fathers. Furthermore, the average difference in the present analysis differs from still another
meta-analysis examining gender differences in childrens
language use: Leaper and Smith (2004) found that girls
were more talkative than boys (d = .11). Whereas the
findings from the present meta-analysis may differ from
these three meta-analyses, they are consistent with the
conclusion reached in two other previous reviews of
gender differences in talkativeness (Hall, 1984; James &
Drakich, 1993). Possible reasons for the discrepancies
between these various reviews will be addressed later
when we discuss the moderator effects.
The results for gender differences in affiliative and
assertive speech were in the expected directions and
extend previous research in these areas. Women were
more likely than men to use affiliative speech and less
likely to use assertive speech, though the effect sizes
were negligible for both. These findings, along with
Leaper et al.s (1998) findings that mothers use more
affiliative speech and less assertive speech than fathers,
provide converging evidence for gender-typed patterns
in affiliative and assertive speech that map onto traditional gender divisions in our society. Gender-typed

353

women may use more affiliative speech to connect with


others, whereas gender-typed men may use more
assertive speech to establish dominance and control
(Aries, 1998; also see Hall, 2006b; Henley, 2001;
LaFrance et al., 2003, for analogous gender differences
in nonverbal behavior). However, as previously noted,
the average differences were negligible in magnitude.
Moreover, as discussed next, both contextual and
methodological factors strongly influenced the extent
and the manner with which women and men were found
to differ in their verbal communication.
Moderators of Gender Differences in Language Use
Consistent with prior meta-analyses of language and
gender (K. J. Anderson & Leaper, 1998; Leaper et al.,
1998; Leaper & Smith, 2004), the present meta-analyses
of adults speech indicated that various factors moderated the likelihood and the magnitude of gender differences in talkativeness, affiliative speech, and assertive
speech. These included the operational definition of the
language construct, various aspects of the interactive
context, and other methodological factors.

Operational Definition
Operational definition was a significant moderator for
all three language constructs. For talkativeness, six specific definitions were contrasted: total words, total turns,
total statements, rate, duration, and MLU or turn. Of
these, there was no average gender difference in measures
of total words spoken. The largest gender differences
favoring men were obtained when measuring MLU, total
statements, or duration. All three of these measures
reflect how much speakers use the conversational floor.
The amount that people talk is often linked with their relative power. Henley (1977, 2001) posited that people in
authority typically talk more than their subordinates. She
then drew a parallel to findings that men talk more than
women in mixed-sex interactionsand argued that this
was a manifestation of male dominance. This proposal is
consistent with Schmid-Masts (2002) meta-analysis indicating a positive average association between speaking
time and trait dominance. (Moreover, the association
between trait dominance and speaking time was stronger
for men than women.)
Total number of turns was the only measure of talkativeness associated with a significant gender difference
favoring women. This may reflect a higher rate of
mutual engagement and support among women participating in a group (e.g., see Tannen, 1994). A related
point is that women may be more likely than men to
observe a norm that one speaker does not hold the floor
too long. Given that the gender effect on total turns was

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

354

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

based on only two samples, the result and our interpretation must be viewed cautiously.
Gender differences in affiliative speech also varied
with the operational definition. General measures of
affiliative speech and two specific measuresactive
understanding and verbal supportwere associated
with significant average effect sizes favoring women.
No gender differences occurred with agreements or
acknowledgments. It is notable that both active understanding and supportive behavior reflect the coordination
of affiliation and self-assertion, whereas agreements and
acknowledgments affirm the other but tend to downplay the self (see Leaper, 1991, 1994, 2000a; Leaper,
Carson, Baker, Holliday, & Myers, 1995; Leaper &
Smith, 2004; Penman, 1980). Thus, relative to men,
women do not typically act in unassertive ways when
demonstrating relatively high levels of affiliation in
their speech.
Variations in effect sizes related to operational definition also occurred with assertive speech. There were no
average differences in informing speech. Perhaps this
should not be a surprise given this is a relatively neutral
speech act (e.g., see Leaper, Tenenbaum, & Shaffer,
1999). Average gender differences were also negligible
for directives and disagreements, which were surprising
given that these were the most power-assertive speech
forms. Whereas men and women did not differ in the
average use of domineering speech, they did differ in
their use of instrumental speech. First, men were more
likely to give suggestions. In addition, the largest average effect size occurred with general task-oriented measures of assertive speech. The present findings suggest a
greater tendency among men than women to approach
social interactions in task-oriented ways. This is consistent with earlier interpretations of the research literature (see Aries, 1996). In contrast, it does not appear
that men were generally domineering in their speech
styles. However, there was still significant heterogeneity
in the effect sizes for both directives and disagreements.
Therefore, gender differences in these highly powerassertive speech forms may depend more on some of the
situational factors reviewed later.
Contrary to prediction, women made significantly
more critical statements than did men, although the
magnitude of the size of the difference was negligible.
Criticism may be part of the socioemotional work in
relationships often associated more with women than
men (see Aries, 1996). For example, researchers studying marital interaction have reported that wives tend to
complain and make demands more than do husbands,
whereas husbands tend to withdraw (Heavey, Layne,
& Christensen, 1993; Noller, 1993). In this light, the
observed pattern may be consistent with certain patterns
of gender typing.

Impact of the Interactive Context


According to socialization explanation, gender differences are a result of men and women having grown
up in different sociocultural contexts. Alternatively,
social constructionists view gender differences as a
result of interactions between men and women in which
differences in status and power are expressed. As
reviewed below, there is support for both approaches,
which is consistent with the view that no single
approach adequately explains gender differences in language use (see Leaper, 2000b; Leaper & Friedman,
2007; Leaper & Smith, 2004).
Undergraduate status. Student status was a significant factor with affiliative and assertive speech. Among
undergraduates, women used more affiliative speech
and less assertive speech than did men. There was no
difference among nonstudents, who were generally
older and more likely to be familiar with one another.
Due to relatively few studies examining either familiar
undergraduates or unfamiliar older adults, it was not
possible to separately test these two factors. We suspect
that self-presentation concerns may diminish with age,
however, as individuals become more comfortable with
their personal identities. (LaFrance et al., 2003, similarly found a decrease in gender differences in smiling
with age.) For that reason, perhaps when being
observed, the younger women were especially concerned with appearing nice (i.e., using affiliative
speech), and the younger men were concerned with
appearing in control (i.e., using assertive speech). As
discussed next, however, the nature of the relationship
between participants rather than age or student status
appeared as a more influential factor.
Relationship between participants. According to contextual and social-role models, gender is most salient
when people are not familiar with one another (Aries,
1996, 1998; Berger & Zelditch, 1998; Deaux & Major,
1987; Eagly, 1987). In the absence of specific status
information, people tend to rely on diffuse status characteristics, such as gender, to guide their behavior (e.g.,
W. Wood & Karten, 1986). This pattern was seen in
each of the three sets of meta-analyses. That is,
womens greater use of affiliative speech, mens greater
talkativeness, and mens greater use of assertive speech
were more likely during interactions between strangers
than close relations. In their meta-analysis, LaFrance
et al. (2003) found an analogous effect of familiarity on
gender differences (favoring women) in smiling.
When specific relationship types were considered,
some interesting subtleties occurred with talkativeness.
First, the largest difference in talkativeness favoring
men occurred during interactions between spouses or

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


partners. The finding suggests that marriage may be a
relationship context where men tend to dominate the
conversational floor. Given the common incidence of
communication problems in dating relationships (e.g.,
Leaper & Anderson, 1997) and marriages (e.g.,
Gottman & Carrere, 1994), future researchers may
want to consider relative talkativeness as a possible
index of relationship satisfaction.
In contrast, studies of two parents with their children
were associated with greater talkativeness among
women than men. In their meta-analysis of parent
gender differences in speech to children, Leaper et al.
(1998) also found more talkativeness among mothers
than fathers. To the extent that child care has traditionally been the domain of women, it is particularly
striking that interacting with children was the one
relationship context in which women were more talkative than men.
Group size. Competition and jockeying for status
tend to be greater in groupsespecially among men
(and boys)than during more intimate contexts (Bales
& Borgatta, 1955; Benenson et al., 2001; Maccoby,
1998). Also, self-presentation concerns are heightened
in larger groups (Deaux & Major, 1987). Contrary to
these notions, however, group size did not significantly
moderate gender differences in assertive speech.
Gender composition. We proposed that gender composition was a key moderator for testing and contrasting the social constructionist or the socialization
interpretations. Confirmation for the social constructionist explanation was expected if effect sizes were significantly larger during mixed-gender than same-gender
interactions (see Carli, 1990; Henley, 2001). That is,
during mixed-gender interactions, the relative status of
the two genders would be relevant. In contrast, support
for the socialization explanation was presumed if
gender differences were greater during same-gender
than mixed-gender interactions (see Carli, 1990). That
is, with same-gender partners, women and men would
be more likely to enact shared gender-typed social
norms.
The social constructionist model was most consistent
with observed gender differences in talkativeness. Men
were more talkative than were women during mixedgender interactions. In contrast, there was no significant
difference during same-gender interactions. This suggests that some men may attempt to dominate the social
interaction with women through holding the conversational floor (see Henley, 2001). Future research should
consider how other factors, such as the relationship
between the conversation partners, affects the likelihood
of this pattern.

355

In contrast, the pattern associated with the socialization interpretation characterized gender differences in
affiliative and assertive speech. First, women were more
likely than men to use affiliative speech during samegender interactions; no significant difference occurred
in mixed-gender interactions. This pattern parallels
those reported in meta-analyses testing for gender differences in self-disclosure (Dindia & Allen, 1992) and
smiling (LaFrance et al., 2003). In the first case, the
authors found that women tended to disclose more than
men, and that the difference was likely in same-gender
but not mixed-gender interactions. The other study
found women smiled more than did men, and that the
effect was larger in same-gender than cross-gender
interactions. These two reports and the present metaanalysis therefore suggest that gender differences in
affiliative or expressive behavior may reflect gendertyped social norms. As developmental research has
highlighted, affiliation is emphasized during girls
socialization in the family (Leaper, 2002; Leaper et al.,
1998) and the peer group (Leaper & Friedman, 2007;
Leaper & Smith, 2004).
The socialization interpretation also fit the findings
for assertive speech. Whereas men used more assertive
speech than women in same-gender pairs, there was
essentially no difference in mixed-gender dyads. This
pattern is surprising in light of prior suggestions that
gender differences in assertive speech are more likely in
mixed-gender than same-gender interactions (e.g., Aries,
1998; Carli, 1990). The latter argument is based on the
premise that men use power-assertive speech to dominate women. However, perhaps men are more concerned with competing for dominance when interacting
with other men (Maltz & Borker, 1982)especially
with either strangers or in larger groups (previously discussed). Indeed, some gender differences in nonverbal
behavior (e.g., gaze, smiling, interpersonal distance) are
more likely in same-gender than mixed-gender interactions (Hall, 1984; LaFrance et al., 2003). Also, during
mixed-gender interactions, perhaps women tend to
accommodate to mens more assertive speech style, and
thereby women and men end up using comparable
amounts of assertive speech (e.g., see Miller, Danaher, &
Forbes, 1986). In this regard, this might be a pattern of
women having learned to play by the mens rules (i.e.,
masculine-stereotyped norms).
Having offered the above interpretations of the
observed gender composition effects, one caveat is worth
noting: Testing for gender effects in same-gender and
mixed-gender groups typically involves different types of
statistical analyses. Tests of gender differences in samegender groups are generally between-group effects,
whereas tests of gender differences within mixed-gender
groups are always within-group effects. Therefore, we

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

356

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

should be cautious when comparing and contrasting the


relative effect sizes from these two conditions.
Researchers presence. The presence of the researcher
during the participants observed interaction was a significant moderator of gender differences in assertive
speech. When the researcher was not present, men
tended to use more assertive speech than did women. In
contrast, if the researcher was present, women were the
ones who used more assertive speech. Perhaps in the
presence of an authority figure (i.e., the researcher),
some men toned down their efforts to be in control and
some women were more comfortable asserting themselves. However, this finding and interpretation need to
be viewed cautiously, because there were only two
samples when the researcher was present.
Setting. As previously noted, gender differences in
social behavior are generally more likely in unfamiliar situations (Deaux & Major, 1987), and that includes laboratory settings. This was seen in all three sets of
meta-analyses. Womens greater use of affiliative speech
and mens greater talkativeness and use of assertive speech
tended to be more likely when the research transpired in a
university lab than another location. In their meta-analysis
of gender differences in smiling, LaFrance et al. (2003)
also found a larger effect size in lab settings. One obvious
implication of this pattern of results is that researchers
may be more likely to elicit gender-typed behavior when
their studies are situated in their research labs.
Activity. Many gender-related behaviors are mediated through the type of activity being pursued. Hence,
when studies are based on unstructured contexts in
which the participants choose their own activities, one
might assume that people select different activities
based on their roles, expectations, and preferences. For
example, studies suggest that women are more likely
than men to discuss socioemotional-oriented topics
with one another, whereas men are more likely to discuss instrumental-oriented topics (Bischoping, 1993).
According to the socialization explanation, larger
gender differences might be anticipated during gendertyped activities wherein women and men are often
expected to enact normative behavior for their gender.
Conversely, based on the social constructionist account,
one would assume that women and men behave similarly when placed in the same situation. Therefore,
gender differences in social behavior may be diminished
when the type of activity is controlled (e.g., LaFrance
et al., 2003; Leaper et al., 1998). In the present metaanalysis, gender differences in talkativeness, affiliative
speech, and assertive speech did vary according to the
activity setting. Moreover, activity was the only factor

to emerge as a significant (or nearly significant) predictor in the exploratory regression analyses for all three
language constructs.
Large effect sizes favoring greater talkativeness
among men were observed in studies in which participants were assigned to discuss nonpersonal topics or
disagreements. Why would men be especially more talkative when specific topics were assigned to discuss? One
interpretation is that many men tended to approach an
assigned topic as an opportunity to lecture about their
views or as a task to accomplish. In contrast, perhaps it
was more likely for women to talk about an assigned
topic in a more relaxed, back-and-forth manner. It is
interesting that there was a negligible average gender
difference when an explicitly instrumental task (deliberation) was assigned. Perhaps these instrumental tasks
involved stronger demand characteristics that affected
mens and womens behavior in a similar manner.
Self-disclosure and child-oriented activities were two
activities in which women were more talkative than men.
Both of these are feminine-stereotyped activities, and therefore it is notable that they were the situations in which
women were more talkative than men. To the extent that
they generally spend time in these activities, women may be
more comfortable than men talking in these contexts. Also,
some men may tend to view these socioemotional activities
as womens work and therefore become less engaged
than do most women (Aries, 1998).
In conjunction with the relative absence of a gender
difference in talkativeness during unstructured activities,
the results indicate that gender differences in talkativeness are not absolute but rather depend on the social context. Thus, there were contexts where either men were
more talkative, women were more talkative, or no average difference was seen. This pattern strongly supports a
social-constructionist interpretation. In other words, variations in talkativeness likely depend on how individuals
interpret particular situations. At the same time, however, how women and men interpret situations likely
depends on their socialization (Leaper, 2000b).
The activity was also an aspect of the context that
moderated gender effects on affiliative speech. Womens
greater use of affiliative speech was limited to instrumental activities that included a combination of deliberations, assigned nonpersonal topics, and other tasks.
There was no average difference in the other settings.
This pattern was not anticipated; our interpretation is
therefore speculative: Perhaps women are more likely
than men to offer verbal support to their partners in the
cooperative pursuit of the exercises. In contrast, men
may be more likely to approach instrumental activities
in a competitive manner, and thereby emphasize control
and emotional restraint (Deaux & Major, 1987;
Miller et al., 1986).

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


Finally, the activity was also a moderator of gender
differences in assertive speech. Mens greater use of
assertive speech was specific to discussions of nonpersonal topics and deliberations. In contrast, there were
negligible differences for other types of activities,
including unstructured activities and disagreements.
There was a sizable effect size favoring women over
men in the use of assertive speech during interactions
with children. These results suggest yet another way in
which the interpersonal context may have moderated
not only the magnitude but also the direction of gender
differences in assertive speech. Thus, in instrumental
activities, men tended to use more assertive speech; in
more interpersonal-oriented contexts, women used
more assertive language.

Other Methodological Moderators


As highlighted in any research methods textbook,
how an investigator measures a construct can influence
the findings that ensue. The present meta-analyses
underscore this point. Besides operational definition
(previously discussed), several methodological factors
moderated the magnitude and the direction of gender
difference in language use. These included two measurement qualities (observation length and recording
method) and four publication characteristics (author
gender, gender focus of study, publication source, and
year of study). Each of these factors is discussed below.
To examine whether observation length was a significant moderator, we contrasted studies in which the
observations were very brief (4-8 min), short (10-15 min),
and somewhat long (20-300 min). If gender differences
in language are due more to peoples self-presentation
concerns, then one might expect that effect sizes would be
more likely in briefer encounters. That is, self-monitoring
of ones image might be more salient at the outset of an
interaction. Indeed, some studies have suggested that
gender differences in social behavior are more likely in
shorter interactions (Eagly & Karau, 1991; Fagot, 1985;
Wheelan & Verdi, 1992). Alternatively, if gender differences in communication reflect underlying stylistic preferences, then one might anticipate larger effect sizes with
longer interactions. As an interaction progresses, peoples
personal styles are more likely to emerge (e.g., Deaux &
Major, 1987). Our results indicated that observation
length was a significant moderator of gender differences
in affiliative speech. A curvilinear pattern occurred
whereby women used more affiliative speech during very
brief observations and somewhat long observations, but
no gender differences were found during intermediate
observations. This pattern suggests that gender differences
may be influenced both by self-presentation concerns (i.e.,
in very brief encounters) and gender-related social norms
(i.e., in longer encounters).

357

The method for recording behavior (video recording,


audio recording, or on-site coding) significantly moderated gender differences in assertive speech. What we
found was that audio and video recordings were significantly different than on-site coding. (Only four studies
employed on-site coding; this result therefore should be
viewed cautiously.) The ability to transcribe and review
observations gives audio and video recordings more
accuracy than on-site coding (Bordens & Abbott, 2002;
Fagot & Hagan, 1988). However, audio-only recordings are potentially subject to biased coding when evaluating assertiveness due to the tonal qualities such as
loudness and voice deepness that may favor men (Hall,
1984).
Next, we turn to the publication characteristics that
were investigated. In the present meta-analysis, there
were no apparent biases toward reporting gender differences based on either the authors gender or the
studys focus on gender. Some prior meta-analyses of
gender differences in language use have reported that
author gender moderated the effects (K. J. Anderson &
Leaper, 1998; Leaper et al., 1998), whereas others have
not (Leaper & Smith, 2004).
Another tested moderator was the publication
source. This factor was not related to average gender
differences in either affiliative or assertive speech.
However, the average effect size for gender differences
in talkativeness (favoring men) was larger in studies
published in top-tier journals than those published in
other sources. This may reflect a bias toward publishing
studies that illustrate a large gender difference rather
than either a small difference or no difference. We do
not consider it a serious confound due to the number of
studies examining gender differences in talkativeness as
well as the inclusion of several studies with no gender
difference (see Table 4).
Finally, we examined whether year of study moderated gender differences. Given the dramatic changes in
womens status and roles in North American society
during the past half century, we anticipated average
gender differences would be smaller in more recent
studies. Other meta-analyses have similarly indicated
some areas in which gender differences have been
declining over the years, such as measures of language
ability (Hyde & Linn, 1988), mathematics performance
(Hyde, Fennema, & Lamon, 1990), gender attitudes
(Twenge, 1997a), and agentic self-concept (Twenge,
1997b). Leaper and Smiths (2004) meta-analysis found
that gender differences in boys and girls talkativeness,
affiliative speech, and assertive speech had declined over
the years. In the present meta-analysis, the cohortrelated effect was seen with affiliative and assertive
speech: In both cases, significant average gender differences occurred among studies published before 1985,

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

358

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

but there was no difference among studies published


since 1985. During the early years of the feminist movement, it was primarily womens gender roles that
changed through the adoption of masculine-stereotyped
roles and behaviors (Twenge, 2001). Perhaps our results
indicate that North American men are increasingly
adopting feminine-stereotyped behaviors through a
greater use of affiliative speech. The trend may be especially likely among college students who represent a
high proportion of the samples in the studies.
General Discussion
With all three language measures, the findings for the
contextual moderators lend support to social constructionist models of gender to the extent that the incidence,
the magnitude, and sometimes the direction of gender difference depended on particular situations. One argument
among some researchers who take a social constructionist approach is that gender differences in communication
are the manifestation of mens dominance over women
(e.g., Henley, 1977, 2001) as well as individuals selfpresentation concerns (e.g., Berger & Zelditch, 1998;
Deaux & Major, 1987; Eagly, 1987). If men are using
language to control women, Carli (1990) reasoned that
support for the dominance explanation would be indicated if gender effects were more likely during mixedgender than same-gender interactions. This was the
observed pattern for gender differences in talkativeness,
but the opposite pattern was found with both affiliative
or assertive speech forms. The results therefore suggest
that gender differences in talkativeness may reflect a tendency among some men to control the conversational
floor when interacting with women. The finding is compatible with the meta-analysis of K. J. Anderson and
Leaper (1998) that found men were more likely than
women to interrupt their conversational partners.
The situational influences revealed in the meta-analyses are also consistent with expectation-states theory
(Berger & Zelditch, 1998) and the contextual-interactive
model of gender (Deaux & Major, 1987). A postulate
in both models is that gender differences in social
behavior are more likely in situations when the participants concerns for self-presentation are heightened
particularly in uncertain or ambiguous social settings.
For example, this includes interactions with strangers.
Overall, gender-typed differences in both affiliative and
assertive speech were more likely during interactions
between strangers than between close relations. In addition, gender differences in all three language constructs
tended to be more likely in research labs than in more
naturalistic settings. (Also, studies in labs tend to
involve strangers.) Given both the unfamiliar nature of
a lab setting as well as possible concerns with appearing

normal, gender may take on more salience in this


context. Thus, women and men may tend to rely on
socialized gender scripts to guide their behaviorbut
mostly in unfamiliar social situations. In these ways,
gender differences in social behavior appear to be less
fixed within the person and more a matter of other
peoples expectations (Berger & Zelditch, 1998; Deaux
& Major, 1987; Eagly et al., 2000). Also, when interacting with same-gender peers, shared gender norms
may additionally contribute to the likelihood of engaging in some forms of gender-typed behavior.
Support for the socialization explanation was also
found. Most notably, this was indicated by the greater
likelihood of gender differences in affiliative and
assertive speech during same- than mixed-gender interactions. The underlying rationale is that group differences are more likely when speakers are interacting with
others who share the same background and social
norms. At the same time, gender differences in all three
language measures varied according to other features in
the interactive context, such as the relationship between
the participants or the type of activity. In these ways,
both individual and situational variables influenced
peoples behavior.
Finally, the pattern of results associated with the metaanalysis of gender effects in language behavior did not
support the strong-effects biological explanation but provided partial confirmation to both the socialization and
the social-constructionist explanations. The strong biological explanation for gender differences in language was
undermined by the negligible overall effect sizes for
gender. The socialization explanation was somewhat supported by the greater talkativeness of women than men
during socioemotional activities and the greater talkativeness of men during the relatively more instrumental activities. However, some researchers might interpret the latter
findings as providing support for the weak-effects version
of the biological explanation. Also, it is possible that some
weak sex-linked biological influences are subject to modification by past experience or situational demands.
In conclusion, we propose the results of the metaanalyses bolster arguments for social rather than strong
biological influences on the incidence and the magnitude of gender differences in language use. The social
constructionist explanation was upheld by the consistent pattern of findings indicating the context-specific
nature of the gender effects. The socialization explanation was supported by findings of possible historical
changes as well as indications that certain gender differences occurred only in same-gender dyads or groups. As
proposed in this article and elsewhere (Leaper, 2000b;
Leaper & Friedman, 2007; Leaper & Smith, 2004;
W. Wood & Eagly, 2002), the social-constructionist and
the socialization approaches may offer complementary

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


explanations for the manifestation and the maintenance
of gender-related variations in social behavior. From
early childhood, dominant cultural practices establish
conditions that often call for different roles and statuses
for girls and boys. These different opportunities foster
corresponding gender differences in expectations, preferences, and competencies. Thus, to best understand the
gender-related variations in social behavior, we will need
to better formulate and take into account a combination
of cultural, institutional, interpersonal, individual, and
biological factors.

NOTES
1. Studies of nonverbal communication point to analogous genderrelated variations (e.g., see Hall, 2006b; Henley, 2001; LaFrance,
Hecht, & Paluck, 2003). However, as Hall (2006a) noted, nonverbal behavior is notoriously ambiguous in meaning (p. 388). In contrast, the meaning of verbal behavior is relatively easier to interpret.
2. Other authors have described affiliative and assertive interpersonal behavior using a variety of terms. Some notable examples
include task-oriented and socioemotional activity (Bales, 1950),
instrumental and expressive behavior (Parsons & Bales, 1955), dominance and love (Leary, 1957), control and affection (Schutz, 1958),
agency and communion (Bakan, 1966), power and affiliation
(McClelland, 1987; Wiggins, 1973), and other-transforming and selftransforming orientations (Selman & Demorest, 1984), respectively.
Although assertion and affiliation are not mutually exclusive psychological acts, they can be viewed as separate dimensions (see Leaper,
1991, 1994, 2000a; Leaper, Tenenbaum, & Shaffer, 1999; Penman,
1980). According to this scheme, a speech act may be both assertive
and affiliative when it is collaborative. Examples would include suggestions for joint activity (Lets go for a walk) or elaborating in relevant ways on the other speakers topic. Speech acts that are assertive
but low in affiliation are controlling (e.g., directives, criticism). Speech
acts that are affiliative but low in assertion are obliging (e.g., agreements). Finally, speech acts that are low in both affiliation and assertion indicate withdrawal.
3. Earlier, it was noted that the effect size for gender differences in
tests of language production was d = .33. Language production was
one type of language measure examined, and it was the one that had
the largest effect size. Other measures included tests of general verbal
ability (d = .20), anagrams (d = .22), essay writing (d = .09), reading
comprehension (d = .03), vocabulary (d = .02), SAT-Verbal (d = .11),
and analogies (d = .16).
4. Total turns may be most useful as an index of talkativeness
when considering a group of individuals. Within a dyad, the number
of turns is necessarily equal (unless one person does a monologue).
Within a larger group, it is possible for group members to vary in taking conversational turns. Accordingly, a follow-up test was carried
out to consider if group size moderated the gender effect among studies using total turns. The contrast was significant, QB(1) = 4.00, p <
.05, with a greater number of turns among women than men in
groups (k = 2; d = .52, p < .05) but not in dyads (k = 2; d = .13, n.s.).

REFERENCES
References marked with an asterisk indicate studies included in the
meta-analyses.
*Adams, K. A. (1980). Who has the final word? Sex, race, and dominance behavior. Journal of Personality & Social Psychology, 38, 1-8.
Andersen, P. A. (2006). The evolution of biological sex differences
in communication. In D. J. Canary & K. Dindia (Eds.), Sex

359

differences and similarities in communication (2nd ed., pp.


117-135). Mahwah, NJ: Lawrence Erlbaum.
Anderson, K. J., & Leaper, C. (1998). Meta-analyses of gender effects
on conversational interruption: Who, what, when, where, and
how. Sex Roles, 39, 225-252.
Anderson, L. R., & Blanchard, P. N. (1982). Sex differences in task
and social-emotional behavior. Basic & Applied Social
Psychology, 3, 109-139.
*Argyle, M., Lalljee, M., & Cook, M. (1968). The effects of visibility
on interaction in a dyad. Human Relations, 21, 3-17.
*Aries, E. (1982). Verbal and nonverbal behavior in single-sex and
mixed-sex groups: Are traditional sex roles changing?
Psychological Reports, 51, 127-134.
Aries, E. (1996). Men and women in interaction: Reconsidering the
difference. New York: Oxford University Press.
Aries, E. (1998). Gender differences in interaction: A reexamination.
In D. J. Canary & K. Dindia (Eds.), Sex differences and similarities in communication: Critical essays and empirical investigations
of sex and gender in interaction (pp. 65-81). Mahwah, NJ:
Lawrence Erlbaum.
*Athenstaedt, U., Haas, E., & Schwab, S. (2004). Gender role selfconcept and gender-typed communication behavior in mixed-sex
and same-sex dyads. Sex Roles, 50, 37-52.
Bakan, D. (1966). The duality of human existence. Boston: Beacon.
Bales, R. F. (1950). Interaction process analysis: A method for the
study of small groups. Reading, MA: Addison-Wesley.
Bales, R. F. (1970). Personality and social behavior. New York: Holt,
Rinehart, & Winston.
Bales, R. F., & Borgatta, E. F. (1955). Size of group as a factor in the
interaction profile. In P. Hare, E. F. Borgatta, & R. F. Bales (Eds.),
Small groups: Studies in social interaction (pp. 495-512). Toronto,
Canada: Random House.
Beall, A. E. (1993). A social constructionist view of gender. In A. E.
Beall & R. J. Sternberg (Eds.), The psychology of gender (pp. 127147). New York: Guilford.
Benenson, J. F., Nicholson, C., Waite, A., Roy, R., & Simpson, A.
(2001). The influence of group size on childrens competitive
behavior. Child Development, 72, 921-928.
Berger, J., & Zelditch, M. (1998). Status, power and legitimacy:
Strategies and theories. New Brunswick, NJ: Transaction.
*Bilous, F. R., & Krauss R. M. (1988). Dominance and accommodation in the conversational behaviors of same- and mixed-gender
dyads. Language and Communication, 8, 183-194.
Bischoping, K. (1993). Gender differences in conversation topics,
1922-1990. Sex Roles, 28, 1-18.
Bordens, K. S., & Abbott, B. B. (2002). Research design and methods:
A process approach (5th ed.). New York: McGraw-Hill.
*Brooks, V. R. (1982). Sex differences in student dominance behavior
in female and male professors classrooms. Sex Roles, 8, 683-690.
*Brouwer, D., Gerritsen, M., & DeHaan, D. (1979). Speech differences between women and men: On the wrong track? Language in
Society, 8, 33-50.
*Brown, P. (1990). Gender, politeness, and confrontation in
Tenejapa. Discourse Processes, 13, 123-141.
*Caldwell, M. A., & Peplau, L. A. (1982). Sex differences in same-sex
friendship. Sex Roles, 8, 721-732.
*Campbell, K. E., Kleim, D. M., & Olson, K. R. (1992). Conversational
activity and interruptions among men and women. Journal of Social
Psychology, 132, 419-421.
*Carli, L. L. (1989). Gender differences in interaction style and influence. Journal of Personality & Social Psychology, 56, 565-576.
*Carli, L. L. (1990). Gender, language, and influence. Journal of
Personality & Social Psychology, 59, 941-951.
*Case, S. S. (1988). Cultural differences, not deficiencies: An analysis of
managerial womens language. In S. Rose & L. Larwood (Eds.),
Womens careers: Pathways and pitfalls (pp. 41-63). New York: Praeger.
*Cashdan, E. (1998). Smiles, speech, and body posture: How women
and men display sociometric status and power. Journal of
Nonverbal Behavior, 22, 209-228.
*Chelune, G. J. (1976). A multidimensional look at sex and target
differences in disclosure. Psychological Reports, 39, 259-263.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

360

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

*Cherulnik, P. D. (1979). Sex differences in the expression of emotion


in a structured social encounter. Sex Roles, 5, 413-424.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
*Craig, D., & Pitts, M. K. (1990). The dynamics of dominance in
tutorial discussions. Linguistics, 28, 125-138.
Crawford, M. E. (1995). Talking difference: On gender and language.
Thousand Oaks, CA: Sage.
*Crosby, F., Jose, P., & Wong-McCarthy, W. (1981). Gender,
androgyny, and conversational assertiveness. In C. Mayo & N. M.
Henley (Eds.), Gender and nonverbal behavior (pp. 151-169).
New York: Springer-Verlag.
*Dabbs J. M., Jr., & Ruback, R. B. (1984). Vocal patterns in male
and female groups. Personality and Social Psychology Bulletin,
10, 518-525.
*Davis, B. M., & Gilbert, L. A. (1989). Effect of dispositional and situational influences on womens dominance expression in mixedsex dyads. Journal of Personality & Social Psychology, 57,
294-300.
Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of gender-related behavior. Psychological Review,
94, 369-389.
Dindia, K., & Allen, M. (1992). Sex differences in self-disclosure: A
meta-analysis. Psychological Bulletin, 112, 106-124.
Dindia, K., & Canary, D. J. (2006). Sex differences and similarities in
communication (2nd ed.). Mahwah, NJ: Lawrence Erlbaum.
*Dixon, J. A., & Foster, D. H. (1998). Gender, social context,
and back channel responses. Journal of Social Psychology, 138,
134-136.
*Dovidio, J. F., Heltman, K., Brown, C. E., Ellyson, S. L., & Keating,
C. F. (1988) Power displays between men and women in discussions of gender-linked tasks: A multicultural study. Journal of
Personality and Social Psychology, 55, 580-587.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role
interpretation. Hillsdale, NJ: Lawrence Erlbaum.
Eagly, A. H., & Carli, L. L. (1981). Sex of researchers and sex-typed
communications as determinants of sex differences in influenceability: A meta-analysis of social influence studies. Psychological
Bulletin, 90, 1-20.
Eagly, A. H., & Karau, S. J. (1991). Gender and the emergence of
leaders: A meta-analysis. Journal of Personality & Social
Psychology, 60, 685-710.
Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory
of sex differences and similarities: A current appraisal. In T. Eckes
& H. M. Trautner (Eds.), The developmental social psychology of
gender (pp. 123-174). Mahwah, NJ: Lawrence Erlbaum.
*Eskilson, A., & Wiley, M. G. (1976). Sex composition and leadership in small groups. Sociometry, 39, 183-194.
*Fagot, B. (1978). The influence of sex of child on parental reactions
to toddler children. Child Development, 49, 459-465.
Fagot, B. I. (1985). A cautionary note: Parents socialization of boys
and girls. Sex Roles, 12, 471-476.
Fagot, B., & Hagan, R. (1988). Is what we see what we get? Comparisons of taped and live observations. Behavioral Assessment,
10, 367-374.
Fitch, R. H., & Bimonte, H. A. (2002). Hormones, brain, and behavior: Putative biological contributions to cognitive sex differences.
In A. McGillicuddy-De Lisi & R. De Lisi (Eds.), Biology, society,
and behavior: The development of sex differences in cognition
(Vol. 21, pp. 55-91). Westport, CT: Ablex.
*Frances, S. J. (1979). Sex differences in nonverbal behavior. Sex
Roles, 5, 519-535.
*Gall, M. D., Hobby, A. K., & Craik, K. H. (1969). Non-linguistic
factors in oral language productivity. Perceptual & Motor Skills,
29, 871-874.
Gleason, J. B., & Ely, R. (2002). Gender differences in language development. In A. M. DeLisi and R. DeLisi (Eds.), Biology, society,
and behavior: The development of sex differences in cognition (pp.
127-154). Westport, CT: Ablex.
Gottman, J. M., & Carrere, S. (1994). Why cant men and women
get along? Developmental roots and marital inequities. In

D. J. Canary & L. Stafford (Eds.), Communication and relational


maintenance (pp. 203-229). San Diego, CA: Academic Press.
Graddol, D., & Swann, J. (1989). Gender voices. Oxford: Basil
Blackwell.
*Grotevant, H. D., & Cooper, C. R. (1985). Patterns of interaction in
family relationships and the development of identity exploration
in adolescence. Child Development, 56, 415-428.
Hall, J. A. (1984). Nonverbal sex differences: Communication accuracy and expressive style. Baltimore: The Johns Hopkins
University Press.
Hall, J. A. (2006a). Nonverbal behavior, status, and gender: How do
we understand their relations? Psychology of Women Quarterly,
30, 384-391.
Hall, J. A. (2006b). Womens and mens nonverbal communication:
Similarities, differences, stereotypes, and origins. In V. Manusov
& M. L. Patterson (Eds.), The Sage handbook of nonverbal communication (pp. 201-218). Newbury Park, CA: Sage.
*Hammen, C., & Peplau, L. A. (1978). Brief encounters: Impact of
gender, sex-role attitudes, and partners gender on interaction and
cognition. Sex Roles, 4, 75-90.
*Hannah, A., & Murachver, T. (1999). Gender and conversational
style as predictors of conversational behavior. Journal of
Language & Social Psychology, 18, 153-174.
*Hauser, S. T., Book, B. K., Houlihan, J., Powers, S. F., Weiss-Perry,
B., Follansbee, D., et al. (1987). Sex differences within the family:
Studies of adolescent and parent family interactions. Journal of
Youth & Adolescence, 16, 199-220.
*Hawkins, K., & Power, C. B. (1999). Gender differences in questions during small decision-making group decisions. Small Group
Research, 30, 235-256.
*Heavey, C. L., Layne, C., & Christensen, A. (1993). Gender and
conflict structure in marital interaction: A replication and extension. Journal of Consulting & Clinical Psychology, 61, 16-27.
*Heiss, J. (1962). Degree of intimacy and male-female interaction.
Sociometry, 25, 197-208.
Henley, N. M. (1977). Body politics: Power, sex, and nonverbal communication. Englewood Cliffs, NJ: Prentice Hall.
Henley, N. M. (2001). Body politics. In A. Branaman (Ed.), Self and
society (pp. 288-297). Malden, MA: Blackwell.
Hines, M. (2004). Brain gender. New York: Oxford University Press.
*Hladik E. G., & Edwards, H. T. (1984). A comparative analysis of
mother-father speech in the naturalistic home environment.
Journal of Psycholinguistic Research, 13, 321-332.
Holmes, J., & Meyerhoff, M. (Eds.). (2003). The handbook of language and gender. Cambridge, MA: Blackwell.
Hyde, J. S., Fennema, E., & Lamon, S. J. (1990). Gender differences
in mathematics performance: A meta-analysis. Psychological
Bulletin, 107, 139-155.
Hyde, J. S., & Linn, M. C. (1988). Gender differences in verbal ability: A meta-analysis. Psychological Bulletin, 104, 53-69.
James, D., & Drakich, J. (1993). Understanding gender differences in
amount of talk: A critical review of the research. In D. Tannen
(Ed.), Gender and conversational interaction (pp. 281-312). New
York: Oxford University Press.
Johnson, B. (1989). DSTAT: Software for the meta-analytic review of
research literature. Hillsdale, NJ: Lawrence Erlbaum.
Johnson, B. (1993). DSTAT 1.10 software for the meta-analytic
review of research literature: Upgrade documentation. Hillsdale,
NJ: Lawrence Erlbaum.
*Johnson, C. (1994). Gender, legitimate authority, and leader-subordinate conversations. American Sociological Review, 59, 122-135.
*Johnson, C., Funk, S. J., & Clay-Warner, J. (1998). Organizational
contexts and conversation patterns. Social Psychology Quarterly,
61, 361-371.
*Jones, E. S., Gallois, C., Callan, V. J., & Barker, M. (1995). Language
and power in an academic context: The effects of status, ethnicity,
and sex. Journal of Language and Social Psychology, 14, 434-461.
*Julien, D., Tremblay, N., Belanger, I., Dube, M., Begin, J., &
Bouthillier, D. (2000). Interaction structure of husbands and
wives disclosure of marital conflict to their respective best friend.
Journal of Family Psychology, 14, 286-303.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


*Kelly, J. A., Wildman, H. E., & Urey, J. R. (1982). Gender and sex
role differences in group decision-making social interactions: A
behavioral analysis. Journal of Applied Social Psychology, 12,
112-127.
*Kimble, C. E., & Musgrove, J. I. (1988). Dominance in arguing
mixed-sex dyads: Visual dominance patterns, talking time, and
speech loudness. Journal of Research in Personality, 22, 1-16.
*Kimble, C. E., Yoshikawa, J. C., & Zehr, H. D. (1981). Vocal and
verbal assertiveness in same-sex and mixed-sex groups. Journal of
Personality & Social Psychology, 40, 1047-1054.
*Kollock, P., Blumstein, P., & Schwartz, P. (1985). Sex and power in
interaction: Conversational privileges and duties. American
Sociological Review, 50, 34-46.
LaFrance, M., Hecht, M. A., & Paluck, E. L. (2003). The contingent
smile: A meta-analysis of sex differences in smiling. Psychological
Bulletin, 129, 305-334.
*Leaper, C. (1987). Agency, communion, and gender as predictors of
communication style and being liked in adult male-female dyads.
Sex Roles, 16, 137-149.
Leaper, C. (1991). Influence and involvement in childrens discourse:
Age, gender, and partner effects. Child Development, 62, 797-811.
Leaper, C. (1994). Exploring the consequences of gender segregation
on social relationships. In C. Leaper (Ed.), Childhood gender segregation: Causes and consequences (New Directions for Child
Development, No. 65, pp. 67-86). San Francisco: Jossey-Bass.
*Leaper, C. (1998). Decision-making processes between friends:
Speaker and partner gender effects. Sex Roles, 39, 125-133.
Leaper, C. (2000a). Gender, affiliation, assertion, and the interactive
context of parent-child play. Developmental Psychology, 36, 381393.
Leaper, C. (2000b). The social construction and socialization of
gender. In P. H. Miller & E. K. Scholnick (Eds.), Toward a
feminist developmental psychology (pp. 127-152). New York:
Routledge.
Leaper, C. (2002). Parenting girls and boys. In M. H. Bornstein (Ed.),
Handbook of parenting: Vol. 1. Children and parenting (2nd ed.,
pp. 189-225). Mahwah, NJ: Lawrence Erlbaum.
Leaper, C., & Anderson, K. J. (1997). Gender development and heterosexual romantic relationships during adolescence. In W.
Damon (Series Ed.) & S. Shulman & W. A. Collins (Issue Eds.),
Romantic relationships in adolescence: Developmental perspectives (New Directions for Child Development, No. 78, pp. 85103). San Francisco: Jossey-Bass.
Leaper, C., Anderson, K. J., & Sanders, P. (1998). Moderators of
gender effects on parents talk to their children: A meta-analysis.
Developmental Psychology, 34, 3-27.
*Leaper, C., Carson, M., Baker, C., Holliday, H., & Myers, S. (1995).
Self-disclosure and listener verbal support in same-gender and
cross-gender friends conversations. Sex Roles, 33, 387-404.
Leaper, C., & Friedman, C. K. (2007). The socialization of gender. In
J. Grusec & P. Hastings (Eds.), Handbook of socialization: Theory
and research (pp. 561-587). New York: Guilford.
*Leaper, C., Hauser, S. T., Kremen, A., Powers, S. I., Jacobson, A.
M., Noam, G. G., et al. (1989). Adolescent-parent interaction in
relation to adolescents gender and ego development pathway: A
longitudinal study. Journal of Early Adolescence, 9, 335-361.
Leaper, C., & Smith, T. E. (2004). A meta-analytic review of gender
variations in childrens talk: Talkativeness, affiliative speech, and
assertive speech. Developmental Psychology, 40, 993-1027.
Leaper, C., Tenenbaum, H. R., & Shaffer, T. G. (1999).
Communication patterns of African-American girls and boys from
low-income, urban backgrounds. Child Development, 70, 14891503.
Leary, T. (1957). Interpersonal diagnosis of personality. New York:
Ronald Press.
*Leffler, A., Gillespie, P. L., & Conaty, J. C. (1982). The effects of
status differentiation on nonverbal behavior. Social Psychology
Quarterly, 45, 153-161.
*Leik, R. K. (1963). Instrumentality and emotionality in family interaction. Sociometry, 26, 131-145.
Lippa, R. A. (2005). Gender, nature, and nurture (2nd ed.). Mahwah,
NJ: Lawrence Erlbaum.

361

Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis.


Thousand Oaks, CA: Sage.
Luxen, M. F. (2005). Gender differences in dominance and affiliation
during a demanding interaction. Journal of Psychology:
Interdisciplinary and Applied, 139, 331-347.
Maccoby, E. E. (1998). The two sexes: Growing up apart, coming
together. Cambridge, MA: Harvard University Press.
Maltz, D. N., & Borker, R. A. (1982). A cultural approach to malefemale miscommunication. In J. J. Gumperz (Ed.), Language and
social identity (pp. 196-216). Cambridge, UK: Cambridge
University.
*Margolin, G., & Wampold, B. E. (1981). Sequential analysis of conflict and accord in distressed and nondistressed partners. Journal
of Consulting & Clinical Psychology, 49, 554-567.
*Markel, N. N., Long, J. F., & Saine, T. J. (1976). Sex effects in controversial interaction: Another look at male dominance. Human
Communication Research, 2, 356-364.
*Marlatt, G. A. (1970). A comparison of vicarious and direct reinforcement control of verbal behavior in an interview setting.
Journal of Personality & Social Psychology, 16, 695-703.
*Martin, J. N., & Craig, R. T. (1983). Selecting linguistic sex differences during initial social interactions of same-sex and mixed-sex
student dyads. The Western Journal of Speech Communication,
47, 16-28.
*Martin, M., Davis, P., & Dancer, J. (1996). Conversations between
older men and women: Turn-taking and topics. Perceptual and
Motor Skills, 83, 1330.
McClelland, D. C. (1987). Human motivation. New York:
Cambridge University Press.
*McLaughlin, A. (1991). The effects of agreement, disagreement,
gender and familiarity on patterns of dyadic interaction. Journal
of Language and Social Psychology, 10, 207-213.
*McLaughlin, M. L., Cody, M. J., Kane, M. L., & Robey, C. S.
(1981). Sex differences in story receipt and story sequencing
behaviors in dyadic conversations. Human Communication
Research, 7, 99-116.
*McMillan, J. R., Clifton, A. K., McGrath, D., & Gale, W. S. (1977).
Womens language: Uncertainty or interpersonal sensitivity and
emotionality? Sex Roles, 3, 545-559.
*McMullen, L. M., Vernon, A. E., & Murton, T. (1995). Divisions of
labor in conversations: Are Fishmans results replicable and generalizable? Journal of Psycholinguistic Research, 24, 255-268.
*Mehl, M. R., & Pennebaker, J. W. (2003). The sounds of social life:
A psychometric analysis of students daily social environments and
natural conversations. Journal of Personality & Social
Psychology, 84, 857-870.
*Mikolic, J. M., Parker, J. C., & Pruitt, D. G. (1997). Escalation in
response to persistent annoyance: Groups versus individuals and
gender effects. Journal of Personality & Social Psychology, 72,
151-163.
Miller, P. M., Danaher, D. L., & Forbes, D. (1986). Sex-related strategies for coping with interpersonal conflict in children aged five and
seven. Developmental Psychology, 22, 543-548.
*Moore, S. F., Shaffer, L., Goodsell, D. A., & Baringoldz, G. (1983).
Gender or situationally determined spoken language differences?
The case of leadership situation. International Journal of Womens
Studies, 6, 44-53.
*Mulac, A. (1989). Mens and womens talk in same-gender and
mixed-gender dyads: Power or polemic? Journal of Language and
Social Psychology, 8, 249-270.
*Mulac, A., Wiemann, J. M., Widenmann, S. J., & Gibson, T. W.
(1988). Male/female language differences and effects in same-sex
and mixed-sex dyads: The gender-linked language effect.
Communication Monographs, 55, 315-335.
*Natale, M., Elliot, E., & Jaffe, J. (1979). Vocal interruptions in
dyadic communication as a function of speech and social anxiety.
Journal of Personality & Social Psychology, 37, 865-878.
*Nemeth, C., Endicott, J., & Wachtler, J. (1976). From the 50s to the
70s: Women in jury deliberations. Sociometry, 39, 293-304.
Noller, P. (1993). Gender and emotional communication in marriage:
Different cultures or differential social power? Journal of
Language & Social Psychology, 12, 137-152.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

362

PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW

Parsons, T., & Bales, R. E. (1955). Family socialization and interaction process. Glencoe, IL: Free Press.
Pellegrini, A. D., & Archer, J. (2005). Sex differences in competitive
and aggressive behavior: A view from sexual selection theory. In
B. J. Ellis & D. F. Bjorklund (Eds.), Origins of the social mind:
Evolutionary psychology and child development (pp. 219-244).
New York: Guilford.
*Pellegrini, A. D., Brody, G. H., & Stoneman, Z. (1987). Childrens
conversational competence with their parents. Discourse Processes,
10, 93-106.
Penman, R. A. (1980). Communication processes and relationships.
London: Academic Press.
*Piliavin, J. A., & Martin, R. R. (1978). The effects of the sex composition of groups on style of social interaction. Sex Roles, 4, 281-296.
*Pillon, A., Degauquier, C., & Duquesne, F. (1992). Males and
females conversational behavior in cross-sex dyads: From gender
differences to gender similarities. Journal of Psycholinguistic
Research, 21, 147-172.
*Porter, N., Geis, F. L., Cooper, E., & Newman, E. (1985).
Androgyny and leadership in mixed-sex groups. Journal of
Personality & Social Psychology, 49, 808-823.
Pushkar, D., Basevitz, P., Arbuckle, T., Nohara-LeClair, M., Lapidus,
S., & Peled, M. (2000). Social behavior and off-target verbosity in
elderly people. Psychology and Aging, 15, 361-374.
*Reid, S. A., Keerie, N., & Palomares, N. A. (2003). Language,
gender salience and social influence. Journal of Language and
Social Psychology, 22, 210-233.
*Resick, P. A., Barr, P. K., Sweet, J. J., Kieffer, D. M., Ruby, N. L.,
& Speigel, D. K. (1981). Perceived and actual discriminators of
conflict from accord in marital communication. American Journal
of Family Therapy, 9, 58-68.
Ridgeway, C. L., & Smith-Lovin, L. (1999). The gender system and
interaction. Annual Review of Sociology, 25, 191-216.
*Robey, E. B., Canary, D. J., & Burggraf, C. S. (1998).
Conversational maintenance behaviors of husbands and wives: An
observational analysis. In D. J. Canary & K. Dinidia (Eds.), Sex
differences and similarities in communication: Critical essays and
empirical investigations of sex and gender in interaction (pp. 373392). Mahwah, NJ: Lawrence Erlbaum.
*Roger, D., & Nesshoever, W. (1987). Individual differences in
dyadic conversational strategies: A further study. British Journal
of Social Psychology, 26, 247-255.
*Roger, D. B., & Schumacher, A. (1983). Effects of individual differences on dyadic conversational strategies. Journal of Personality
and Social Psychology, 45, 700-705.
*Roopnarine, J. L., & Adams, G. R. (1987). The interactional teaching patterns of mothers and fathers with their popular, moderately
popular, or unpopular children. Journal of Abnormal Child
Psychology, 15, 125-136.
*Rosenfeld, H. (1966). Approval-seeking and approval-inducing
functions of verbal and nonverbal responses in the dyad. Journal
of Personality & Social Psychology, 4, 597-605.
Roter, D. L., Hall, J. A., & Aoki, Y. (2002). Physician gender effects
in medical communication: A meta-analytic review. JAMA:
Journal of the American Medical Association, 288, 756-764.
*Ruback, R. B., Dabbs, J. M., Jr., & Hopper, C. H. (1984). The
process of brainstorming: An analysis with individual and group
vocal parameters. Journal of Personality & Social Psychology, 47,
538-567.
*Rubin, D. L., & Nelson, M. W. (1983). Multiple determinants of
stigmatized speech style: Womens language, powerless language,
or everyones language. Language & Speech, 26, 273-290.
Ruble, D. N., Martin, C. L., & Berenbaum, S. A. (2006). Gender development. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.),
Handbook of child psychology: Vol. 3. Social, emotional, and personality development (6th ed., pp. 858-932). Hoboken, NJ: John Wiley.
*Sayers, S. L., & Baucom, D. H. (1991). Role of femininity and masculinity in distressed couples communication. Journal of
Personality & Social Psychology, 61, 641-647.
*Schmid-Mast, M. (2001). Gender differences and similarities in
dominance hierarchies in same-gender groups based on speaking
time. Sex Roles, 44, 537-556.

Schmid-Mast, M. (2002). Dominance expressed and inferred through


speaking time: A meta-analysis. Human Communication
Research, 28, 420-450.
Schutz, W. C. (1958). FIRO: A three-dimensional theory of interpersonal behavior. New York: Holt, Rinehart, & Winston.
*Scudder, J. N., & Andrews, P. H. (1995). A comparison of two
alternative models of powerful speech: The impact of power and
gender upon the use of threats. Communication Research Reports,
12, 25-33.
Selman, R. L., & Demorest, A. P. (1984). Observing troubled
childrens interpersonal negotiation strategies: Implications of and
for a developmental model. Child Development, 55, 288-304.
Sharpe, D. (1997). Of apples and oranges, file drawers and garbage:
Why validity issues in meta-analysis will not go away. Clinical
Psychology Review, 17, 881-901.
*Shaw, M. E., & Sadler, O. W. (1965). Interaction patterns in heterosexual dyads varying in degree of intimacy. Journal of Social
Psychology, 66, 345-351.
*Simkins-Bullock, J. A., & Wildman, B. G. (1991). An investigation
into the relationships between gender and language. Sex Roles, 24,
149-160.
*Stets, J. E., & Burke, P. J. (1996). Gender, control, and interaction.
Social Psychology Quarterly, 59, 193-220.
*Stiles, W. B., Lyall, L. M., Knight, D. P., Ickes, W., Waung, M., Hall,
C. L., et al. (1997). Gender differences in verbal presumptuousness
and attentiveness. Society for Personality and Social Psychology,
7, 759-772.
*Stiles, W. B., Walz, N. C., Schroeder, M. A. B., Williams, L. L., &
Ickes, W. (1996). Attractiveness and disclosure in initial encounters of mixed-sex dyads. Journal of Social and Personal
Relationships, 13, 303-312.
*Stuckey, M. F., McGhee, P. E., & Bell, N. J. (1982). Parent-child
interaction:
The
influence
of
maternal
employment.
Developmental Psychology, 18, 635-644.
Sutton, A. J., Duval, S. J., Tweedie, R. L., Abrams, K. R., & Jones,
D. R. (2000). Empirical assessment of effect of publication bias on
meta-analyses. British Medical Journal, 320, 1574-1577.
Swann, J. (1992). Girls, boys, and language. Cambridge, MA:
Blackwell.
Talbot, M. M. (1998). Language and gender. Malden, MA:
Blackwell.
Tannen, D. (1990). You just dont understand: Women and men in
conversation. New York: Morrow.
Tannen, D. (1994). Interpreting interruption in conversation. In
D. Tannen (Ed.), Gender and discourse (pp. 53-84). New York:
Oxford.
Thorne, B. (1993). Gender play: Girls and boys in school. New
Brunswick, NJ: Rutgers University Press.
Thorne, B., & Henley, N. (1975). Language and sex: Difference and
dominance. Rowley, MA: Newbury House.
*Turner, G. S., Tjaden, K., & Weismer, G. (1995). The influence of
speaking rate on vowel space and speech intelligibility for individuals with amyotrophic lateral sclerosis. Journal of Speech &
Hearing Research, 38, 1001-1013.
Twenge, J. M. (1997a). Attitudes toward women, 1970-1995: A
meta-analysis. Psychology of Women Quarterly, 21, 35-51.
Twenge, J. M. (1997b). Changes in masculine and feminine traits over
time: A meta-analysis. Sex Roles, 36, 305-325.
Twenge, J. M. (2001). Changes in womens assertiveness in response
to status and roles: A cross-temporal meta-analysis, 1931-1993.
Journal of Personality and Social Psychology, 81, 133-145.
West, C., & Zimmerman, D. H. (1985). Gender, language, and discourse. In T. A. van Dijk (Ed.), Handbook of discourse analysis:
Vol. 4. Discourse analysis in society (pp. 103-124). London:
Academic Press.
Wheelan, S. A., & Verdi, A. F. (1992). Differences in male and female
patterns of communication in groups: A methodological artifact?
Sex Roles, 27, 1-15.
Wiggins, J. S. (1973). Personality and prediction. Reading, MA:
Addison-Wesley.
Wood, J. T. (2007). Gendered lives: Communication, gender, and
culture (7th ed.). Belmont, CA: Wadsworth.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014

Leaper, Ayres / GENDER AND LANGUAGE


*Wood, M. (1966). The influence of sex and knowledge of communication effectiveness on spontaneous speech. Word, 22,
112-137.
Wood, W., & Eagly, A. H. (2002). A cross-cultural analysis of the
behavior of women and men: Implications for the origins of sex
differences. Psychological Bulletin, 128, 699-727.

363

*Wood, W., & Karten, S. J. (1986). Sex differences in interaction style


as a product of perceived sex differences in competence. Journal of
Personality & Social Psychology, 50, 341-347.
*Yamada, E. M., Tjosvold, D., & Draguns, J. G. (1983). Effects of
sex-linked situations and sex composition on cooperation and
style of interaction. Sex Roles, 9, 541-553.

Downloaded from psr.sagepub.com at Universiti Teknologi MARA (UiTM) on November 19, 2014